WO2019037014A1 - 一种图像检测的方法、装置及终端 - Google Patents

一种图像检测的方法、装置及终端 Download PDF

Info

Publication number
WO2019037014A1
WO2019037014A1 PCT/CN2017/098787 CN2017098787W WO2019037014A1 WO 2019037014 A1 WO2019037014 A1 WO 2019037014A1 CN 2017098787 W CN2017098787 W CN 2017098787W WO 2019037014 A1 WO2019037014 A1 WO 2019037014A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
detected
feature
detection
color
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2017/098787
Other languages
English (en)
French (fr)
Inventor
丁欣
董辰
卢恒惠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to EP17922433.2A priority Critical patent/EP3664016B1/en
Priority to PCT/CN2017/098787 priority patent/WO2019037014A1/zh
Priority to CN201780064306.5A priority patent/CN109844804B/zh
Priority to US16/640,945 priority patent/US11321830B2/en
Publication of WO2019037014A1 publication Critical patent/WO2019037014A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2200/00—Indexing scheme for image data processing or generation, in general
    • G06T2200/24—Indexing scheme for image data processing or generation, in general involving graphical user interfaces [GUIs]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20221—Image fusion; Image merging
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30041—Eye; Retina; Ophthalmic
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30088—Skin; Dermal

Definitions

  • Ways to detect skin health include: detecting according to images or detecting the skin with a sensor.
  • the image quality of the existing photographing technology is very similar to that used in professional skin detecting equipment.
  • the big gap, the clarity of the facial skin detail features often can not meet the requirements of detection accuracy, so that the user's advice is not accurate enough.
  • an embodiment of the present invention provides a method for image detection.
  • the method specifically includes: determining, by the terminal, the original image to be detected according to the feature to be detected in the skin detection mode; the terminal processing the original image to be processed according to a preset rule corresponding to the feature to be detected to obtain a detection specific image; the terminal determining the regular image; The terminal determines an image to be displayed according to the detection result image and the regular image.
  • a detection-dedicated image is obtained, the detection result image is determined according to the detection-dedicated image, and the detection result image and the regular image are processed to obtain an image to be displayed. . It solves the problem that the imaging quality does not meet the detection accuracy of health detection, and it can not provide users with convenient, accurate and professional skin health detection and evaluation anytime, anywhere.
  • the feature to be detected includes: at least one of a skin color, a color spot, a red zone, an acne, an oil component, a wrinkle, an eye feature, and a pore
  • the eye feature may include: an eye bag At least one of dark circles and fine lines of the eyes.
  • the original image to be detected includes at least one of a color to-be-detected original image and a black-and-white to-be-detected original image.
  • the method for determining, by the terminal in the skin detection mode, the original image to be detected according to the feature to be detected may include: when the feature to be detected includes: skin color, color patch, red zone, and acne In at least one of the original images to be detected is a color to be detected original image; when the feature to be detected includes at least one of oil, wrinkles, eye features and pores, the original image to be detected is a black and white image to be detected.
  • different original images to be detected may be provided for the characteristics of the feature types to be detected, and an unprocessed original image with higher precision is provided for post processing. To ensure the high precision of the original image to be detected.
  • the method for “terminal determining a normal image” may include: the terminal fuses at least two sub-images of the regular image, wherein the at least two sub-images include: a color regular sub-image and a black-and-white regular sub-image At least one of them.
  • the method for the “processing of the original image to be detected by the terminal according to the preset rule corresponding to the feature to be detected” to obtain the detection-specific image may include:
  • the preset rule corresponding to the skin color includes at least one sub-rule of illumination equalization, white balance adjustment, and image smoothing.
  • preset rules corresponding to at least one of the color spot and the red zone include: contrast transformation, image sharpening, and red-brown X color space transformation (red At least one sub-rule in brown X, RBX).
  • the preset rules corresponding to the acne include at least one sub-rule of contrast transformation, image sharpening, and hue-saturation-value (HSV).
  • different preset rules are provided for processing, for example, after the processing according to the preset rule corresponding to the skin color, the skin noise can be reduced; according to preset rules corresponding to at least one of the color spot and the red zone. After processing, you can increase the contrast and increase the sharpness of the image to the preset value, and then highlight the color spots through the B channel (brown channel) of the RBX space.
  • the R (red channel) channel can highlight the red area, according to the pre-corresponding acne
  • the rule processing can increase the contrast and increase the sharpness of the image to a preset value, and then highlight the acne feature through the HSV spatial image; according to the preset rule processing corresponding to the oil component, the oil component can be detected through the skin reflective portion, by increasing The contrast to the preset value can further highlight the reflective portion, and the oil component is detected according to the reflective portion; according to the preset rule processing corresponding to at least one of the eye feature and the pore, the feature region can be highlighted by increasing the contrast and increasing the sharpness of the image. .
  • the terminal displays the image to be displayed.
  • the method can be classified according to the features to be detected, and can be classified into a group of skin color, color spots, red areas and acne, and a group of oil, wrinkles, eye features and pores, which will be classified according to classification.
  • the feature to be detected selects a suitable camera, and according to the selected camera, an original image to be detected (ie, an image of an unprocessed RAW format) that is more suitable for the feature to be detected is captured.
  • the camera can be selected in the skin detection mode, and the shooting mode can be selected.
  • the feature to be detected in the shooting mode can be selected by the user, which is not limited herein.
  • the terminal automatically detects all the features to be detected by default, so the camera determined by the terminal may also be a dual camera, and the dual camera may include: a color camera and a black and white camera.
  • the first determining unit processes and detects the original file to be detected in the skin detecting mode to obtain a detection-dedicated image
  • the second determining unit determines the detection result image according to the detection-dedicated image, and then the detection result image and the conventional image.
  • the image is processed to obtain an image to be displayed. It solves the problem that the imaging quality does not meet the detection accuracy of health detection, and it can not provide users with convenient, accurate and professional skin health detection and evaluation anytime, anywhere.
  • the feature to be detected includes at least one of skin color, color spots, red areas, acne, oil, wrinkles, eye features, and pores, wherein the eye features may include: an eye bag At least one of dark circles and fine lines of the eyes.
  • the original image to be detected when the feature to be detected includes at least one of a skin color, a color spot, a red zone, and a acne, the original image to be detected is a color image to be detected; and the feature to be detected includes: When at least one of oil, wrinkles, eye features, and pores is present, the original image to be detected is a black-and-white original image to be detected.
  • different original images to be detected may be provided for the characteristics of the feature types to be detected, and an unprocessed original image with higher precision is provided for post processing. To ensure the high precision of the original image to be detected.
  • the foregoing “second determining unit” may be configured to fuse at least two sub-images of the regular image, wherein the at least two sub-images include: at least one of a color regular sub-image and a black-and-white regular sub-image One.
  • the preset rules corresponding to the skin color include:
  • the preset rule corresponding to the skin color includes at least one sub-rule of illumination equalization, white balance adjustment, and image smoothing.
  • the preset rule corresponding to at least one of the color spot and the red zone includes at least one of a contrast transform, an image sharpening, and an RBX color space transform. rule.
  • the preset rules for acne include: contrast transformation, image sharpening, and at least one sub-rule in the HSV spatial image.
  • the preset rule corresponding to the oil component includes: contrast transformation.
  • At least one of the preset rules corresponding to the eye feature and the pore includes at least one sub-rule of contrast transformation, illumination equalization, and image sharpening.
  • the foregoing “matching unit” is further configured to process the detection result image and the regular image by using image matching and image fusion to determine an image to be displayed.
  • the foregoing apparatus further includes: a display module, configured to display an image to be displayed.
  • the foregoing “first determining unit” may be further configured to: determine a feature to be detected according to the feature to be detected in the skin detecting mode; and determine, in the skin detecting mode, the image to be detected according to the feature to be detected.
  • the camera of the image the camera includes at least one of a color camera and a black and white camera.
  • different preset rules are provided for processing, for example, after the processing according to the preset rule corresponding to the skin color, the skin noise can be reduced; according to preset rules corresponding to at least one of the color spot and the red zone.
  • the preset corresponding to the acne Rule processing can increase the contrast and increase the sharpness of the image to the preset value, and then highlight the acne feature through the HSV space image
  • the preset rules of the oil component the oil detection can be completed through the skin reflection part, by increasing the contrast
  • the reflective portion can be further highlighted, and the oil component is detected according to the reflective portion; according to the preset rule processing corresponding to at least one of the eye feature and the pore, the feature region can be highlighted by increasing the contrast and increasing the sharpness of the
  • the method may include: determining, in the skin detection mode, that the camera that captures the original image to be detected is colored if the feature to be detected includes at least one of skin color, color patch, red zone, and acne a camera; if the feature to be detected includes at least one of oil, wrinkles, eye features, and pores, determining, in the skin detection mode, the camera that captures the original image to be detected is a black and white camera.
  • an embodiment of the present invention provides a terminal.
  • the terminal includes: a processor, configured to determine an original image to be detected according to a feature to be detected in a skin detection mode; the processor is further configured to: perform, according to a preset rule corresponding to the feature to be detected, the original image to be detected Processing the detection-dedicated image; and determining the regular image; the processor is further configured to: detect the detection-dedicated image, and determine the detection result image; the processor is further configured to determine the image to be displayed according to the detection result image and the regular image.
  • the processor detects and detects the original file in the skin detection mode, and obtains the detection specific image, and the processor can further determine the detection result image according to the detection specific image, and then perform the detection result image and the regular image.
  • the image is obtained by the processing. It solves the problem that the imaging quality does not meet the detection accuracy of health detection, and it can not provide users with convenient, accurate and professional skin health detection and evaluation anytime, anywhere.
  • the feature to be detected includes at least one of a skin color, a color spot, a red zone, a acne, an oil component, a wrinkle, an eye feature, and a pore
  • the eye feature may include: At least one of an eye bag, a dark eye, and an eyelet.
  • the “original image to be detected” includes: at least one of a color image to be detected and an original image to be detected in black and white.
  • the original image to be detected when the feature to be detected includes at least one of skin color, color patch, red zone, and acne, the original image to be detected is a color image to be detected; and the feature to be detected includes: oil component When at least one of wrinkles, eye features, and pores, the original image to be detected is a black-and-white original image to be detected. Because the types of the features to be detected are different, in an alternative implementation, different original images to be detected may be provided for the characteristics of the feature types to be detected, and an unprocessed original image with higher precision is provided for post processing. To ensure the high precision of the original image to be detected.
  • the foregoing “processor” may be configured to fuse at least two sub-images of a regular image, wherein at least two sub-images include at least one of a color regular sub-image and a black-and-white regular sub-image.
  • the method may include:
  • the preset rule corresponding to the skin color includes at least one sub-rule of illumination equalization, white balance adjustment, and image smoothing.
  • a preset rule corresponding to at least one of the color patch and the red region includes at least one of contrast transformation, image sharpening, and RBX color space transformation. Sub-rules.
  • the preset rule corresponding to the oil component includes: contrast transformation.
  • At least one of the preset rules corresponding to the eye feature and the pore includes: at least one sub-rule of contrast transformation, illumination equalization, and image sharpening .
  • the adjustment range of the “white balance adjustment” is a preset value
  • the transformation space of the “RBX color space conversion” and the “HSV spatial image” is a preset value.
  • the foregoing “processor” may be further configured to: process the detection result image and the regular image by using image matching and image fusion to determine an image to be displayed.
  • the foregoing “processor” may be further configured to: determine a feature to be detected in a skin detection mode; and determine, in the skin detection mode, a camera that captures an original image to be detected according to the feature to be detected, where the camera includes : at least one of a color camera and a black and white camera.
  • the processor determines, in the skin detection mode, that the camera that captures the original image to be detected is a color camera. If the feature to be detected includes at least one of oil, wrinkles, eye features, and pores, the processor determines, in the skin detection mode, that the camera that captures the original image to be detected is a black and white camera.
  • the foregoing “processor” includes: an image signal processor and a central processing unit, and the image signal processor is configured to determine an original image to be detected according to the feature to be detected in the skin detection mode; The preset rule corresponding to the feature to be detected, the original image to be detected is processed to obtain a detection-specific image; and the regular image is determined.
  • an embodiment of the present application provides a computer program product, comprising computer readable instructions, when a computer reads and executes the computer readable instructions, such that the computer performs the foregoing first aspect and its optional implementation. method.
  • FIG. 2 is a schematic structural diagram of an image detecting terminal according to an embodiment of the present invention.
  • FIG. 4( a ) is a to-be-detected original image in a processed RAW format according to an embodiment of the present invention
  • FIG. 4(b) is an image of a JPEG format according to an embodiment of the present invention.
  • FIG. 5(b) is an interface diagram of another mode selection according to an embodiment of the present invention.
  • FIG. 6(a) is a schematic diagram of an original image to be detected containing acne according to an embodiment of the present invention
  • FIG. 6(b) is a schematic diagram of a special image for detecting acne according to an embodiment of the present invention.
  • FIG. 7(a) is a schematic diagram of an image with a large oil content according to an embodiment of the present invention.
  • FIG. 7(b) is a schematic diagram of an image with normal oil content according to an embodiment of the present invention.
  • FIG. 8 is a schematic diagram of an image of a detection result of acne detection according to an embodiment of the present invention.
  • FIG. 9 is a schematic diagram of an image to be displayed containing a color patch according to an embodiment of the present invention.
  • FIG. 10 is a schematic diagram of an image to be displayed including an eye feature according to an embodiment of the present invention.
  • FIG. 12 is a diagram showing a display interface including an image to be displayed according to an embodiment of the present invention.
  • FIG. 16 is a flowchart of still another method for image detection according to an embodiment of the present invention.
  • FIG. 17 is a schematic diagram of an apparatus for image detection according to an embodiment of the present invention.
  • FIG. 18 is a schematic diagram of a terminal according to an embodiment of the present invention.
  • the embodiment of the invention provides a method, a device and a terminal for image detection, which are processed and detected by detecting an original file in a skin detection mode to obtain a detection result image, and then matching and merging the detection result image with a regular image.
  • FIG. 1 is a flow chart of a method for processing an image by an image signal processor.
  • an image service processor ISP
  • the image signal processor can be disposed in the terminal.
  • the ISP receives the unprocessed image in the format of RAW, that is, the original image to be detected appearing in the following, and performs preprocessing as follows: performing hot pixel correction on the original image to be detected in the format RAW, Demoscal, noise reduction, shading correction, geometric correction, color correction, tone curve adjustment, and edge enhancement ), obtaining at least one image output in the YUV format.
  • the ISP sends the preprocessed image to a JPEG encoder, and the JPEG encoder compresses the preprocessed image and outputs the image in JPEG format.
  • the JPEG image obtained by the method for processing an unprocessed image cannot satisfy the accuracy requirement of the subsequent skin feature detection, and the specific performance is as follows: First, in S102, the unprocessed image in the format RAW is damaged. The image obtained in JPEG format is compressed, and some skin detail features become inconspicuous after being compressed by the JPEG encoder, which makes the subsequent detection of the JPEG format image more difficult, resulting in inaccurate results.
  • the JPEG image retains only 8 bits of color information per pixel, and the color information of the RAW format is 10-16 bits, which indicates that JPEG is caused.
  • the color gradation of the image is greatly reduced, and the adjustable range of contrast, light, etc. is also reduced correspondingly, resulting in inaccurate results.
  • the white balance adjustment range of the JPEG image obtained after the processing of this method is limited, and the color temperature adjustment usually loses the detailed information, resulting in inaccurate subsequent results.
  • the ISP can also directly output the raw image of the format RAW directly.
  • FIG. 2 is a schematic structural diagram of an image detecting terminal according to an embodiment of the present invention.
  • the terminal may include a mobile phone, a tablet computer, a notebook computer, a personal digital assistant (PDA), a point of sales (POS), and a vehicle-mounted computer.
  • the terminal may include at least a camera 210, a processing device, an image signal processor 230, a display 240, and a communication bus 250.
  • the camera 210, the processing device, the image signal processor (ISP) 230, and the display 240 communicate with each other through the communication bus 250. Connect and complete communication with each other.
  • ISP image signal processor
  • the image signal processor 230 is configured to call an operation instruction and receive an image of the unprocessed RAW format, that is, the original image captured by the camera 210 and perform the following operations:
  • the original image to be detected is processed according to a preset rule corresponding to the feature to be detected to obtain a detection specific image
  • the image to be displayed is determined based on the detection result image and the regular image.
  • the above method may be processed in the image signal processor 230. It is also possible to perform processing in the image signal processor 230 and the processing device respectively, for example, “determining the original image to be detected according to the feature to be detected in the skin detection mode; and performing the original image to be detected according to the preset rule corresponding to the feature to be detected.
  • the step of processing the detection-dedicated image; and determining the regular image is processed in the image signal processor 230, and then "detecting the detection-dedicated image, determining the detection result image; determining the image to be detected based on the detection result image and the regular image"
  • the steps are processed in the processing device.
  • the above steps can be understood as a parallel method added in the image signal processor 230, that is, the step of acquiring the detection dedicated image does not take processing by a processor other than the image signal processor 230.
  • the process can be done at image signal processor 230.
  • the processor will call the conventional image saved in the terminal memory 233 and the detection-dedicated image, and then the processor detects the image again.
  • the dedicated image is processed to generate a detection result image.
  • the above operation is implemented in the image signal processor 230 as a preferred solution. The reason is as follows.
  • the above operation is implemented in the image signal processor 230, and no additional RAW image is stored, and the occupied space is small. . Secondly, it is implemented in the image signal processor 230, which is low in cost and fast in processing speed, and improves the user experience.
  • the conventional image obtained at this step is generated simultaneously with the generation of the detection-dedicated image.
  • the processing device in the device may further include: a processor and a memory.
  • the processor may be a central processing unit (CPU), and the processor may be another general-purpose processor, a digital signal processor (DSP), or an application specific integrated circuit (ASIC). , a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, and the like.
  • the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
  • the processor can also receive the detection-dedicated image determined by the image signal processor 230 and the regular image, detect and determine the detection result image for the detection-dedicated image, and the processor matches and fuses the detection result image with the regular image to determine the image to be displayed.
  • the memory may include a read only memory and a random access memory, and in particular, a detection specific image may be stored and instructions and data may be supplied to the image signal processor 230.
  • a portion of the memory may also include a non-volatile random access memory.
  • the memory can also store information of the device type.
  • the communication bus 250 may include a power bus, a control bus, a status signal bus, and the like in addition to the data bus. However, for clarity of description, various buses are labeled as communication bus 250 in the figure.
  • the display 240 is used to display an image to be displayed.
  • the terminal may further include a Bluetooth, an antenna, a microphone, and the like, and details are not described herein again. It will be understood by those skilled in the art that the terminal structure shown in FIG. 2 does not constitute a limitation to the terminal, and may include more or less components than those illustrated, or a combination of certain components, or different component arrangements.
  • the executor of the method may be the terminal shown in FIG. 2.
  • the following embodiments take a mobile phone as an example, and the embodiment is combined with FIG. 3 to FIG. Detailed Description.
  • FIG. 3 is a flowchart of a method for image detection according to an embodiment of the present invention.
  • the method 300 can detect the skin of the whole body.
  • the following method takes the facial skin as an example.
  • the method 300 can specifically include the following steps:
  • S310 The terminal determines an original image to be detected according to the feature to be detected in the skin detection mode.
  • the terminal detects the original image by using a color camera in the skin detecting mode, and then the terminal Determining to detect an original image to be detected in the original image, the original image to be detected is a color image to be detected.
  • the user selects a skin detection mode on the terminal, wherein the mode selection in FIG. 5(a) may include multiple modes, such as a normal mode, a skin detection mode, and other modes.
  • the skin detecting mode may include a plurality of skin detecting modes such as a first skin detecting mode and a second skin detecting mode.
  • the first skin detecting mode may include at least one of skin color, color patch, red zone and acne;
  • the second skin detecting mode may include features to be detected: oil, wrinkles, and eye features. And at least one of the pores.
  • the terminal uses a dual camera to capture an original image to be detected (ie, color) in the skin detection mode.
  • the camera and the black and white camera simultaneously take a picture), and then the terminal extracts only the image of the unprocessed RAW format corresponding to the color camera as the original image to be detected.
  • S320 The terminal processes the original image to be detected according to a preset rule corresponding to the feature to be detected to obtain a detection specific image.
  • the format of the detection-dedicated image may be a JPEG format.
  • the step is a detection-specific image generated after the terminal directly processes the original image to be detected, and then generates the detection-specific image directly after the processing according to the preset rule. Instead of first obtaining the original image to be detected, the original image to be detected is stored, and then post-processing is performed using a preset rule corresponding to the feature to be detected to generate a detection-dedicated image.
  • the preset rule includes a plurality of sub-rules. It is to be understood that the terminal determines a preset rule corresponding to the to-be-detected feature according to the to-be-detected feature, and obtains a detection-specific image after the processing corresponding to the to-be-detected feature after the preset rule processing corresponding to the detection feature, the preset
  • the rules specifically include the following:
  • a preset rule corresponding to at least one of the color patch and the red zone includes at least one sub-rule of contrast transform, image sharpening, and RBX color space transform.
  • the color spot can be highlighted by increasing the contrast and increasing the sharpness of the image to a preset value, and then passing through the B channel (brown channel) of the RBX space.
  • the R (red channel) channel can highlight the red zone.
  • the preset rules corresponding to the oil component include: contrast transformation.
  • contrast transformation For example, in order to clearly distinguish between oily skin and normal skin, the present embodiment provides an image as shown in Fig. 7, and Fig. 7(a) shows a skin having a large oil content, and a place having a bright portion is a large oil.
  • the oil component can be detected through the reflective part of the skin, and by increasing the contrast to a preset value, the reflective portion can be further highlighted, and the oil component can be detected according to the reflective portion.
  • the preset rule corresponding to at least one of the eye feature and the pore includes at least one sub-rule of contrast transformation, illumination equalization, and image sharpening.
  • the feature area can be highlighted by increasing the contrast and increasing the degree of image sharpening according to preset rules corresponding to at least one of the eye features and the pores.
  • the adjustment range of the white balance adjustment is a preset value
  • the transformation space of the RBX color space conversion and the HSV spatial image is a preset value.
  • the above-mentioned detection-dedicated image may be a plurality of images, and the image format of the detection-dedicated image may be a JPEG format.
  • the preset value needs to be set in advance based on a large amount of experimental data.
  • S330 The terminal determines a regular image.
  • the conventional image can be obtained directly by conventional pre-processing, or can be obtained by fusing at least two sub-images of a regular image.
  • the at least two sub-images include at least one of a color regular sub-image and a black-and-white regular sub-image, and the format of the regular image may be a JPEG format, and the format of the sub-image may be a JPEG format.
  • the step of merging the plurality of sub-images of the regular image may include separately performing pixel-level synchronization on the at least two sub-images, then aligning, and finally extracting the dominant parts in the at least two image data by using an image fusion algorithm, and synthesizing an image.
  • the dominant portion may include: a color portion in the color regular sub-image and a detail portion of the shadow in the black and white regular sub-image.
  • S340 The terminal detects the detection specific image and determines the detection result image.
  • the skin-color detection-specific image is detected by using a skin-color detection algorithm corresponding to the skin color
  • the skin-color detection algorithm may include: capturing the acquired skin color detection according to the preset standard image The image is color-calibrated, and then the skin-in-color detection-dedicated image is compared with a standard skin color number to determine the closest skin color, and the detection result image containing the skin color feature mark is determined.
  • the spot-and-red color corresponding to at least one of the color spot and the red zone is utilized.
  • At least one detection algorithm in the region detects at least one detection specific image in the stain and the red region, and the at least one detection algorithm in the stain and the red region may include: at least one of the stain and the red region A detection specific image, extracting an R channel (or B channel) component in the RBX image, performing image processing (such as edge detection) on the component, and determining detection of at least one of the color patch and the red region Result image.
  • the acne detection algorithm corresponding to the acne detection algorithm is used to detect the acne detection specific image, and the acne detection algorithm may include: a dedicated image for acne detection, that is, the above-mentioned HSV spatial image is made. Image processing to determine an image of the test result containing the acne mark.
  • FIG. 8 is a detection result image of acne detection, and the detection result image may include: a mark of a feature to be detected (the rectangular frame shown in FIG. 8 is a mark, and the shape of the mark may be more than a rectangle The shape is not limited here).
  • the oil-inspection detection algorithm is used to detect the oil-inspection-specific image, and the oil-splitting detection algorithm may include: detecting, by the image processing means, the brightness in the oil-inspection-specific image is large The area is calculated by occupying the area of the area to determine the oil size, and the image of the detection result containing the oil characteristic mark is determined.
  • the detection-dedicated image corresponding to at least one of the eye feature and the pore is at least one of the eye feature and the pore-detection-dedicated image
  • the eye feature corresponding to at least one of the eye feature and the pore is utilized
  • At least one detection algorithm of the pores detects at least one of the ocular features and the pore-specific detection images, and determines a detection result image including at least one of the ocular features and the pores, the ocular features and the pores
  • At least one of the detection algorithms may include: an image processing type algorithm, such as edge detection, image filtering, or the like, or a machine learning type method.
  • S350 The terminal determines an image to be displayed according to the detection result image and the regular image.
  • the terminal processes the at least one detection result image and the regular image by using image matching and image fusion to obtain the image to be displayed.
  • the image matching is to identify a point of the same name between the at least one detection result image and the regular image by the matching algorithm. For example, in the image matching, by comparing the correlation coefficient between the target area in the two or more images and the window of the same size in the search area, the center point of the window corresponding to the largest correlation coefficient in the search area is taken as the same name point.
  • Image fusion is to perform pixel level synchronization on at least one detection result image and a regular image respectively, and then align, finally extract the dominant part in at least two image data by image fusion algorithm, and combine one image to extract the respective channels to the maximum extent.
  • the favorable information is finally integrated into high-quality images to improve the utilization of image information, improve the spatial resolution and spectral resolution of the original image, and facilitate detection.
  • the above image fusion process is similar to the fusion in the S330 step, except that the original images of the two are different, and the methods adopted are all performed by using an image fusion algorithm.
  • the detection result image may be an image including a flag to be detected; that is, an image formed by extracting an image of the detection-dedicated image containing the feature portion to be detected, and performing the image to be detected.
  • the mark is specifically as shown by the rectangular box in FIG.
  • the image to be displayed may be an image obtained by matching the image containing the feature to be detected and the regular image, as shown in FIG. 9 to FIG.
  • the area corresponding to the detection result image in the regular image is matched with the detection result image, and the area corresponding to the detection result image in the regular image is overwritten/replaced using the detection result image, thereby obtaining the area to be displayed.
  • the image to be displayed may be as shown in FIG.
  • FIG. 9 may be an image to be displayed corresponding to the color spot, and 901 may be displayed as the position and size of the color spot.
  • the preset rule and the detection algorithm corresponding to the color spot may make the position display of the color spot more obvious.
  • FIG. 10 may be an image to be displayed of an eye feature, and 1001 may be displayed as the position and size of the eye feature.
  • the preset rules and detection algorithms corresponding to the eye feature may highlight the problem of the eye feature.
  • 11 can be an image to be displayed of the pores, and 1101 can be displayed as the position and size of the pores.
  • the preset rules and detection algorithms corresponding to the pores can highlight the position of the pores for the user to view.
  • the image format of the image to be displayed may be a JPEG image.
  • S360 The terminal displays an image to be displayed.
  • the content displayed by the terminal may include an image to be displayed, and may also include at least one of an image to be displayed and a text description and care information obtained according to the image to be displayed.
  • the form of the nursing information is various.
  • the form of the nursing information may be searching for a preset expert suggestion library, and then obtaining a nursing suggestion corresponding to the detection result image; or may be pushing the image according to the detection result.
  • Corresponding links and the like for example, an expert blog or a website
  • the online consultants may also be selected to provide face-to-face online care suggestions according to the image to be detected, which is not limited in the embodiment of the present invention.
  • the content displayed by the terminal may further include: information such as ranking or rating of the detection result obtained by the detection result image in the same age, and it is necessary to explain that the ranking or the rating
  • the subsequent evaluations are based on the detection results, and no limitation is imposed on the embodiments of the present invention.
  • the display interface may include only the image to be displayed 1201, and the image to be displayed may include information such as a detection result image and a care suggestion. It may also be an image to be displayed as shown in FIG. 13 , the image to be displayed includes at least two parts: a detection result image 1301 and a suggestion 1302 , which may include: detection results, care suggestions, skin health status of the same age person Rankings and ratings.
  • the display image is not limited in any way, and the recommended portion is not limited.
  • the suggestion portion includes the 1201 portion and the recommended portion of 1302.
  • FIG. 14 is a flowchart of a method for image detection according to an embodiment of the present invention.
  • the terminal can activate the camera shooting function to perform shooting in the skin detection mode provided by the skin health detection application. It is also possible to directly activate the camera capture function of the terminal and then select the skin detection mode for shooting.
  • the user can select a feature to be detected in the skin health detection mode, such as: skin color.
  • the terminal determines that the feature to be detected is the skin color, and the terminal invokes the camera to perform the shooting, and the terminal acquires the original data of the comprehensive captured image, wherein the camera includes: a color camera and/or a black and white camera.
  • the terminal reads the RAW color to-be-detected original image 1401 in the original data, and processes the color to-be-detected original image 1401 by using the preset rule 1 corresponding to the skin color to obtain the detection-dedicated image 1402.
  • the preset rule 1 corresponding to the skin color may include at least one sub-rule of illumination equalization, white balance adjustment, and image smoothing. For example, after processing according to the preset rule corresponding to skin color, skin noise can be reduced.
  • the terminal matches and fuses the regular image sub-image 1404 and the regular image sub-image 1504 with the detection result image 1403.
  • An image 1406 to be displayed is obtained.
  • the conventional image sub-image 1404 is obtained by the terminal processing the color to-be-detected original image 1401 by using a general pre-processing, and the color to-be-detected original image 1401 can be captured by a color camera.
  • the conventional image sub-image 1504 is obtained by the terminal processing the black-and-white detection original image 1501 by using a general pre-processing, and the black-and-white original image 1501 can be captured by a black-and-white camera.
  • the terminal displays an image 1406 to be displayed.
  • the image to be displayed 1406 may include: the content displayed by the terminal may include an image to be displayed 1406, and may also include at least one of an image to be displayed 1406 and a text description and care information obtained according to the image 1406 to be displayed.
  • the form of the nursing information is various.
  • the form of the nursing information may be searching for a preset expert suggestion library, and then obtaining a nursing suggestion corresponding to the detection result image; or may be pushing the image according to the detection result.
  • Corresponding links and the like for example, an expert blog or a website
  • the online consultants may also be selected to provide face-to-face online care suggestions according to the image to be detected, which is not limited in the embodiment of the present invention.
  • FIG. 15 is a flowchart of another method for image detection according to an embodiment of the present invention. As shown in FIG. 15, if the user selects at least one of oil, wrinkles, eye features, and pores, the image is detected using the method shown in FIG.
  • the terminal when the user opens the skin health detection application, the terminal can activate the camera shooting function to perform shooting in the skin detection mode provided by the skin health detection application. It is also possible to directly activate the camera capture function of the terminal and then select the skin detection mode for shooting.
  • the oil component can be detected through the reflective part of the skin, and by increasing the contrast to a preset value, the reflective portion can be further highlighted, according to The reflective part detects the oil.
  • the terminal uses the detection algorithm 2 corresponding to the oil component to process the detection-dedicated image 1502 to obtain a detection result image 1503.
  • the detection algorithm 2 corresponding to the oil component may include: detecting an area with a large brightness in the oil detection special image by an image processing means, calculating an area of the area occupying the area of the face to determine the oil size, and taking the calculated oil size as a waiting
  • the oil mark of the dedicated image 1502 is detected to obtain the detection result image 1503, and the detection result is
  • the fruit image 1503 includes: marker information of the oil component, and the marker information of the oil component may include a position of the oil component over the entire face, an oil content, and the like.
  • the terminal matches and fuses the regular image sub-image 1504, the regular image sub-image 1404, and the detection result image 1503.
  • An image 1506 to be displayed is obtained.
  • the conventional image sub-image 1504 is obtained by the terminal processing the black and white to-be-detected original image 1501 by using a general pre-processing, and the black-and-white to-be-detected original image 1501 can be captured by a black-and-white camera.
  • the conventional image sub-image 1404 is obtained by the terminal processing the color detection original image 1401 by using a general preprocessing, and the color to-be-detected original image 1401 can be captured by a color camera.
  • the terminal displays an image 1506 to be displayed.
  • the content displayed by the terminal may further include: information such as ranking or rating of the detection result obtained by the detection result image in the same age, and it is necessary to explain that the ranking or the rating All follow-up evaluations are based on the test results.
  • FIG. 16 is a flowchart of still another method for image detection according to an embodiment of the present invention. As shown in FIG. 16, if the user selects at least one of skin color, stain, red area, and acne and at least one of oil, wrinkles, eye features, and pores, the image is imaged as shown in FIG. Test.
  • the terminal can activate the camera shooting function to perform shooting in the skin detection mode provided by the skin health detection application. It is also possible to directly activate the camera capture function of the terminal and then select the skin detection mode for shooting.
  • the user can select the mode to be detected in the skin health detection mode, such as skin tone and oil.
  • the terminal selects two features to be detected in the skin detection mode, the skin color and the oil component. After selecting the feature to be detected, the terminal calls the color camera to capture the color to be detected original image 1401 and the black and white camera to capture the black and white image to be detected 1501.
  • the image format of the color to-be-detected original image 1401 and the black-and-white to-be-detected original image 1501 may be a RAW format, and the process is performed on the format of the RAW image, and the result obtained by the process is better than that in the compressed JPEG format.
  • the results after image processing are more accurate.
  • the skin detection mode may be a camera shooting function after the skin health detection application is turned on. It is also possible to directly call the camera application and select the skin detection mode for shooting.
  • the terminal processes the color to-be-detected original image according to the preset rule 1 corresponding to the skin color, and obtains the detection-dedicated image 1402 including the skin color.
  • the preset rule 1 corresponding to the skin color may include at least one sub-rule of illumination equalization, white balance adjustment, and image smoothing. For example, after processing according to the preset rule corresponding to skin color, skin noise can be reduced.
  • the terminal detects the detection-dedicated image 1402 including the skin color using the detection algorithm 1 corresponding to the skin color, and obtains the detection result image 1403 including the skin color.
  • the detection algorithm 1 corresponding to the skin color may include: performing color calibration on the acquired skin color detection dedicated image 1402 according to a preset standard picture, and then comparing the skin color detection dedicated image 1402 with a standard skin color number to determine the closest skin color. .
  • the terminal processes the black and white to-be-detected original image 1501 according to the preset rule 2 corresponding to the oil component, and obtains the detection-dedicated image 1502 including the oil component.
  • the preset rule 2 corresponding to the oil component may include: contrast transformation.
  • the oil component can be detected through the skin reflective portion, and by increasing the contrast to a preset value, the reflective portion can be further highlighted, and the oil component can be detected according to the reflective portion.
  • the terminal detects the detection-dedicated image 1502 including the oil component using the detection algorithm 2 corresponding to the oil component, and obtains a detection result image 1503 including the oil component.
  • the detection algorithm 2 corresponding to the oil component may include: detecting, by the image processing means, a region having a large brightness in the oil component detection dedicated image 1502, and calculating an area of the region occupying the face region to determine the oil size.
  • the terminal processes the black and white image to be detected 1501 by using a conventional preprocessing method to obtain a normal image sub image 1504.
  • the image format of the regular image sub image 1504 may be a JPEG format.
  • the terminal fuses the regular image sub-image 1404 and the regular image sub-image 1504 to obtain a regular image.
  • the terminal matches and fuses the above-described detection result image 1401 including the skin color, the detection result image 1501 including the oil component, and the regular image 1605, to obtain an image 1606 to be displayed.
  • FIG. 17 is a schematic diagram of an apparatus for image detection according to an embodiment of the present invention.
  • an embodiment of the present invention provides an apparatus for image detection.
  • the device includes:
  • the first determining unit 1701 is configured to determine an original image to be detected according to the feature to be detected in the skin detecting mode.
  • the detecting unit 1703 is configured to detect the detection-dedicated image and determine the detection result image.
  • the matching unit 1704 is configured to determine an image to be displayed according to the detection result image and the regular image.
  • the above-mentioned features to be detected include at least one of skin color, color spots, red areas, acne, oil, wrinkles, eye features, and pores, wherein the eye features may include: eye bags, dark circles, and fine lines of the eyes. At least one of them.
  • the first determining unit 1701 is further configured to: the original image to be detected includes at least one of a color to-be-detected original image and a black-and-white to-be-detected original image.
  • the original image to be detected is a color image to be detected; and the feature to be detected includes: oil, wrinkles, and eye features And at least one of the pores, the original image to be detected is a black and white image to be detected.
  • different original images to be detected may be provided for the characteristics of the feature types to be detected, and an unprocessed original image with higher precision is provided for post processing. To ensure the high precision of the original image to be detected.
  • the preset rule corresponding to at least one of the color spot and the red zone includes at least one of a contrast transform, an image sharpening, and an RBX color space transform. rule.
  • the preset rules for acne include: contrast transformation, image sharpening, and HSV null At least one sub-rule in the inter-image.
  • the preset rule corresponding to the oil component includes: contrast transformation.
  • At least one of the preset rules corresponding to the eye feature and the pore includes at least one sub-rule of contrast transformation, illumination equalization, and image sharpening.
  • the matching unit 1704 is further configured to process the detection result image and the regular image by using a method of image matching and image fusion to determine an image to be displayed.
  • the first determining unit 1701 is further configured to: determine a feature to be detected according to the feature to be detected in the skin detecting mode; determine, in the skin detecting mode, a camera that captures an original image to be detected according to the feature to be detected, the camera includes: a color camera and a black and white camera At least one of them.
  • different preset rules are provided for processing, for example, after the processing according to the preset rule corresponding to the skin color, the skin noise can be reduced; according to preset rules corresponding to at least one of the color spot and the red zone.
  • the preset corresponding to the acne Rule processing can increase the contrast and increase the sharpness of the image to the preset value, and then highlight the acne feature through the HSV space image
  • the preset rules of the oil component the oil detection can be completed through the skin reflection part, by increasing the contrast
  • the reflective portion can be further highlighted, and the oil component is detected according to the reflective portion; according to the preset rule processing corresponding to at least one of the eye feature and the pore, the feature region can be highlighted by increasing the contrast and increasing the sharpness of the
  • the feature to be detected includes at least one of a skin color, a color spot, a red zone, and a acne, determining, in the skin detection mode, that the camera that captures the original image to be detected is a color camera; if the feature to be detected includes: oil, wrinkles, and eyes At least one of the feature and the pore determines that the camera that takes the original image to be detected is a black and white camera in the skin detection mode.
  • FIG. 18 is a schematic diagram of a terminal according to an embodiment of the present invention.
  • an embodiment of the present invention provides a terminal.
  • the terminal includes: a processor 1801, configured to determine an original image to be detected according to a feature to be detected in a skin detection mode; and the processor is further configured to: process the original image to be detected according to a preset rule corresponding to the feature to be detected And determining a regular image; the processor is further configured to: detect the detection-dedicated image, and determine the detection result image; the processor is further configured to determine the image to be displayed according to the detection result image and the regular image.
  • the processor detects and detects the original file in the skin detection mode, and obtains the detection specific image, and the processor can further determine the detection result image according to the detection specific image, and then perform the detection result image and the regular image.
  • the image is obtained by the processing. It solves the problem that the imaging quality does not meet the detection accuracy of health detection, and it can not provide users with convenient, accurate and professional skin health detection and evaluation anytime, anywhere.
  • the feature to be detected includes at least one of skin color, color spots, red areas, acne, oil, wrinkles, eye features, and pores, wherein the eye features may include: eye bags, dark circles, and fine lines in the eyes. At least one of them.
  • the original image to be detected includes at least one of a color to-be-detected original image and a black-and-white to-be-detected original image.
  • the original image to be detected is a color image to be detected; when the feature to be detected includes: oil, wrinkles, eye features, and At least one of the pores
  • the original image to be detected is black and white, the original image is to be detected.
  • different original images to be detected may be provided for the characteristics of the feature types to be detected, and an unprocessed original image with higher precision is provided for post processing. To ensure the high precision of the original image to be detected.
  • the processor may be further configured to fuse at least two sub-images of the regular image, wherein the at least two sub-images comprise at least one of a color regular sub-image and a black-and-white regular sub-image.
  • the preset rule corresponding to the skin color includes at least one sub-rule of illumination equalization, white balance adjustment, and image smoothing.
  • a preset rule corresponding to at least one of the color patch and the red region includes at least one of contrast transformation, image sharpening, and RBX color space transformation. Sub-rules.
  • the preset rules corresponding to the acne include: contrast transformation, image sharpening, and at least one sub-rule in the HSV spatial image.
  • the preset rule corresponding to the oil component includes: contrast transformation.
  • At least one of the preset rules corresponding to the eye feature and the pore includes: at least one sub-rule of contrast transformation, illumination equalization, and image sharpening .
  • the adjustment range of the white balance adjustment is a preset value
  • the transformation space of the RBX color space conversion and the HSV spatial image is a preset value
  • the processor may be configured to process the detection result image and the regular image by using image matching and image fusion to determine an image to be displayed.
  • the processor can be configured to display the detection feature option in the skin detection mode, and determine the feature to be detected according to the user selection.
  • the processor may be further configured to: determine a feature to be detected in the skin detection mode; determine, in the skin detection mode, a camera that captures an original image to be detected according to the feature to be detected, the camera includes at least one of a color camera and a black and white camera.
  • different preset rules are provided for processing, for example, after the processing according to the preset rule corresponding to the skin color, the skin noise can be reduced; according to preset rules corresponding to at least one of the color spot and the red zone.
  • the preset corresponding to the acne Rule processing can increase the contrast and increase the sharpness of the image to the preset value, and then highlight the acne feature through the HSV space image
  • the preset rules of the oil component the oil detection can be completed through the skin reflection part, by increasing the contrast
  • the reflective portion can be further highlighted, and the oil component is detected according to the reflective portion; according to the preset rule processing corresponding to at least one of the eye feature and the pore, the feature region can be highlighted by increasing the contrast and increasing the sharpness of the
  • the processor determines, in the skin detection mode, that the camera that captures the original image to be detected is a color camera; if the feature to be detected includes: At least one of oil, wrinkles, eye features, and pores, the processor determines in the skin detection mode that the camera that is to take the original image to be detected is a black and white camera.
  • the processor may include: an image signal processor and a central processing unit.
  • An image signal processor configured to determine an original image to be detected according to a feature to be detected in a skin detection mode; The preset rule corresponding to the feature is detected, and the original image to be detected is processed to obtain a detection-specific image; and the regular image is determined.
  • the central processing unit is configured to detect the detection specific image, determine the detection result image, and determine the image to be displayed according to the detection result image and the regular image.
  • the preferred solution of the solution includes: the above processor preferably an image signal processor.
  • the above processor preferably an image signal processor.
  • it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof.
  • software it may be implemented in whole or in part in the form of a computer program product.
  • the computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in accordance with embodiments of the present invention are generated in whole or in part.
  • the computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable device.

Landscapes

  • Engineering & Computer Science (AREA)
  • Quality & Reliability (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)
  • Measuring And Recording Apparatus For Diagnosis (AREA)

Abstract

本发明实施例涉及一种图像检测的方法、装置及终端，涉及多媒体技术领域，具体方法包括：终端在皮肤检测模式下根据待检测特征确定待检测原图像；根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；确定常规图像；对检测专用图像进行检测，确定检测结果图像；根据检测结果图像和常规图像，确定待显示图像。本方案中，通过在皮肤检测模式下对待检测原文件进行处理和检测，得到检测专用图像，根据该检测专用图像确定检测结果图像，再将该检测结果图像和常规图像进行匹配和融合得到了待显示图像。解决了成像质量不满足健康检测的检测精度的要求，以及不能为用户随时随地提供便捷、准确和专业的皮肤健康检测和评估的问题。

Description

一种图像检测的方法、装置及终端 技术领域
本发明实施例涉及多媒体技术领域，尤其涉及一种图像检测的方法、装置及终端。
背景技术
爱美之心，人皆有之。随着生活水平的稳步提升，人们对美的追求也越来越强烈。而皮肤健康是追求美、崇尚美的人士所关注的重点。皮肤健康检测是指对皮肤状况的整体评估，其中包括对肤质、肤色、皱纹细纹、色斑、红区、痤疮、毛孔、黑眼圈和眼袋等的检测和评估。根据这些评估结果，人们可以针对性地采取皮肤护理措施，保持皮肤健康和美丽。
传统的皮肤健康检测通常在医院或美容院由专业人员采用专业设备进行检测，这种方式虽然准确、专业，却缺乏便利性、实时性并且价格昂贵。当前市场上已经出现一些手持式的检测设备，但仍存在携带不便、不普及以及费用高昂等问题。所以，基于终端(例如：手机或电脑等)的皮肤健康检测可以为用户提供便捷、专业的服务，用户只需简单地进行拍照，即可获知自己当前的皮肤状况，并得到相应的护理建议。
检测皮肤健康的方式包括：根据图像进行检测或者利用传感器对皮肤进行检测，然而，在根据图像进行检测的方法中，由于现有拍照技术的成像质量与专业皮肤检测设备中所采用的照片有很大差距，人脸皮肤细节特征的清晰度往往不能满足检测精度的要求，以至于用户得到的建议不够准确。
发明内容
本发明实施例提供了一种图像检测的方法、装置及终端，用以解决成像质量不满足皮肤健康检测的检测精度的要求，以及不能为用户随时随地提供便捷、准确和专业的皮肤健康检测和评估的问题。
第一方面，本发明实施例提供了一种图像检测的方法。该方法具体包括:终端在皮肤检测模式下根据待检测特征确定待检测原图像；终端根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；该终端确定常规图像；该终端根据检测结果图像和常规图像，确定待显示图像。
本方案中，通过在皮肤检测模式下对待检测原文件进行处理和检测，得到检测专用图像，根据该检测专用图像确定检测结果图像，再对该检测结果图像和常规图像进行处理得到了待显示图像。解决了成像质量不满足健康检测的检测精度的要求，以及不能为用户随时随地提供便捷、准确和专业的皮肤健康检测和评估的问题。
在一个可选的实现方式中，上述待检测特征包括：肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种，其中，该眼部特征可以包括：眼袋、黑眼圈和眼部细纹中的至少一个。
在另一个可选的实现方式中，所述待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
在又一个可选的实现方式中，上述“终端在皮肤检测模式下根据待检测特征确定待检测原图像”的方法可以包括：当待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，待检测原图像为彩色待检测原图像；当待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种时，待检测原图像为黑白待检测原图像。
由于该待检测特征的种类不同，所以在可选的实现方式中可以针对该待检测特征种类的特性提供不同的待检测原图像，为后期处理提供一个精度较高的未经处理过的原图像，保证待检测原图像的高精度。
在再一个可选的实现方式中，上述“终端确定常规图像”的方法可以包括：终端融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。
在再一个可选的实现方式中，上述“终端根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像”的方法可以包括：
当待检测特征为肤色时，该肤色对应的预置规则，包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。
当待检测特征为色斑和红区中的至少一种时，该色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和红棕X颜色空间变换(red brown X，RBX)中的至少一个子规则。
当待检测特征为痤疮时，该痤疮对应的预置规则包括：对比度变换、图像锐化和色调-饱和度-明度空间图像(hue-saturation-value，HSV)中的至少一个子规则。
当待检测特征为油分时，该油分对应的预置规则包括：对比度变换。
当待检测特征为眼部特征和毛孔中的至少一种时，该眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
在再一个可选的实现方式中，白平衡调整的调整幅度为预设值，RBX颜色空间变换和HSV空间图像的变换空间为预设值。
根据不同的待检测特征，提供了不同的预置规则进行处理，例如：根据肤色对应的预设规则处理之后，可以减少皮肤噪点；根据色斑和红区中的至少一种对应的预置规则处理之后，可以通过增大对比度和提高图像锐化程度至预设值，再经过RBX空间的B通道(brown通道)凸显色斑，R(red通道)通道可以凸显红区，根据痤疮对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再通过HSV空间图像凸显痤疮特征；根据油分对应的预置规则处理，可以通过皮肤反光部分完成油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分；根据眼部特征和毛孔中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度突出特征区域。
在再一个可选的实现方式中，上述“终端根据检测结果图像和常规图像，确定待显示图像”的方法可以包括：终端利用图像匹配和图像融合的方法对检测结果图像和常规图像进行处理，确定待显示图像。
在再一个可选的实现方式中，终端在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。由于用户在实际操作时，有些待检测原图中的皮肤区域中的待检测 特征是不需要检测的，所以该实现方式中还可以为用户提供选择，该终端在皮肤检测模式下显示检测特征选项，用户可以自主选择，提高用户体验。
在再一个可选的实现方式中，该终端显示该待显示图像。
在再一个可选的实现方式中，上述“终端在皮肤检测模式下根据待检测特征确定待检测原图像”的方法可以包括：终端在皮肤检测模式下根据待检测特征确定拍摄待检测原图像的摄像头，摄像头包括：彩色摄像头和黑白摄像头中的至少一个。
在再一个可选的实现方式中，上述“终端在皮肤检测模式下根据待检测特征确定拍摄待检测原图像的摄像头”的方法可以包括：若待检测特征包括：肤色、色斑、红区和痤疮中的至少一种，则终端在皮肤检测模式下确定拍摄待检测原图像的摄像头为彩色摄像头；若待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种，则终端在皮肤检测模式下确定拍摄待检测原图像的摄像头为黑白摄像头。
为了得到更清晰的检测专用图像，该方法可以根据待检测特征进行分类，可以分类成肤色、色斑、红区和痤疮一组以及油分、皱纹、眼部特征和毛孔一组，将根据分类过的待检测特征选择适合的摄像头，根据选择后的摄像头拍摄更适合待检测特征的待检测原图像(即未经处理的RAW格式的图像)。该摄像头可以是在皮肤检测模式下调用的，该拍摄模式可以选择，拍摄模式的待检测特征可以用户选择，在此不做限定。当用户不对待检测特征进行选择时，该终端自动默认为对待检测特征进行全部检测，所以该终端确定的摄像头也可以为双摄像头，该双摄像头可以包括：彩色摄像头和黑白摄像头。
第二方面，本发明实施例提供了一种图像检测的装置。该装置包括：第一确定单元，用于在皮肤检测模式下根据待检测特征确定待检测原图像；第二确定单元，用于根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；以及确定常规图像；检测单元，用于对检测专用图像进行检测，确定检测结果图像；匹配单元，用于根据检测结果图像和常规图像，确定待显示图像。
本方案中，通过第一确定单元在皮肤检测模式下对待检测原文件进行处理和检测，得到检测专用图像，第二确定单元再根据检测专用图像确定检测结果图像，再对该检测结果图像和常规图像进行处理得到了待显示图像。解决了成像质量不满足健康检测的检测精度的要求，以及不能为用户随时随地提供便捷、准确和专业的皮肤健康检测和评估的问题。
在一个可选的实施方式中，上述待检测特征包括：肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种，其中，该眼部特征可以包括：眼袋、黑眼圈和眼部细纹中的至少一个。
在另一个可选的实现方式中，上述待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
在又一个可选的实现方式中，上述当待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，待检测原图像为彩色待检测原图像；当待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种时，待检测原图像为黑白待检测原图像。
由于该待检测特征的种类不同，所以在可选的实现方式中可以针对该待检测特征种类的特性提供不同的待检测原图像，为后期处理提供一个精度较高的未经处理过的原图像，保证待检测原图像的高精度。
在再一个可选的实现方式中，上述“第二确定单元”可以用于，融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。在再一个可选的实现方式中，肤色对应的预置规则，包括：
当待检测特征为肤色时，该肤色对应的预置规则包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。
当待检测特征为色斑和红区中的至少一种时，色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则。
当待检测特征为痤疮时，痤疮对应的预置规则包括：对比度变换、图像锐化和HSV空间图像中的至少一个子规则。
当待检测特征为油分时，油分对应的预置规则包括：对比度变换。
当待检测特征为眼部特征和毛孔中的至少一种时，眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
在再一个可选的实现方式中，上述“白平衡调整”的调整幅度为预设值，上述“RBX颜色空间变换”和上述“HSV空间图像”的变换空间为预设值。
在再一个可选的实现方式中，上述“匹配单元”还用于，利用图像匹配和图像融合的方法对检测结果图像和常规图像进行处理，确定待显示图像。
在再一个可选的实现方式中，上述“第一确定模块”还用于，在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。
在再一个可选的实现方式中，上述装置还包括：显示模块，用于显示待显示图像。
在再一个可选的实现方式中，上述“第一确定单元”还可以用于：在皮肤检测模式下根据待检测特征确定待检测特征；在皮肤检测模式下根据待检测特征确定拍摄待检测原图像的摄像头，摄像头包括：彩色摄像头和黑白摄像头中的至少一个。
根据不同的待检测特征，提供了不同的预置规则进行处理，例如：根据肤色对应的预设规则处理之后，可以减少皮肤噪点；根据色斑和红区中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再经过RBX空间的B通道(brown通道)凸显色斑，R(red通道)通道可以凸显红区，根据痤疮对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再通过HSV空间图像凸显痤疮特征；根据油分对应的预置规则处理，可以通过皮肤反光部分完成油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分；根据眼部特征和毛孔中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度突出特征区域。
在再一个可选的实现方式中，可以包括：若待检测特征包括：肤色、色斑、红区和痤疮中的至少一种，则在皮肤检测模式下确定拍摄待检测原图像的摄像头为彩色摄像头；若待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种，则在皮肤检测模式下确定拍摄待检测原图像的摄像头为黑白摄像头。
第三方面，本发明实施例提供了一种终端。该终端包括：处理器，用于在皮肤检测模式下根据待检测特征确定待检测原图像；处理器还用于，根据所述待检测特征对应的预置规则，对所述待检测原图像进行处理得到检测专用图像；；以及确定常规图像；处理器还用于，对检测专用图像进行检测，确定检测结果图像；处理器还用于，根据检测结果图像和常规图像，确定待显示图像。
本方案中，通过处理器在皮肤检测模式下对待检测原文件进行处理和检测，得到检测专用图像，处理器还可以再根据检测专用图像确定检测结果图像，再对该检测结果图像和常规图像进行处理得到了待显示图像。解决了成像质量不满足健康检测的检测精度的要求，以及不能为用户随时随地提供便捷、准确和专业的皮肤健康检测和评估的问题。
在一个可选的实施方式中，上述待检测特征包括：肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种，其中，该该眼部特征可以包括：眼袋、黑眼圈和眼部细纹中的至少一个。
在另一个可选的实现方式中，上述“待检测原图像”包括：待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
在又一个可选的实现方式中，当待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，待检测原图像为彩色待检测原图像；当待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种时，待检测原图像为黑白待检测原图像。由于该待检测特征的种类不同，所以在可选的实现方式中可以针对该待检测特征种类的特性提供不同的待检测原图像，为后期处理提供一个精度较高的未经处理过的原图像，保证待检测原图像的高精度。
在再一个可选的实现方式中，上述“处理器”可以用于，融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。在再一个可选的实现方式中，可以包括：
当待检测特征为肤色时，该肤色对应的预置规则包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。
当待检测特征为色斑和红区中的至少一种时，该色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则。
当待检测特征为痤疮时，该痤疮对应的预置规则包括：对比度变换、图像锐化和HSV空间图像中的至少一个子规则。
当待检测特征为油分时，该油分对应的预置规则包括：对比度变换。
当待检测特征为眼部特征和毛孔中的至少一种时，该眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
在再一个可选的实现方式中，上述“白平衡调整”的调整幅度为预设值，上述“RBX颜色空间变换”和上述“HSV空间图像”的变换空间为预设值。
在再一个可选的实现方式中，上述“处理器”还可以用于：利用图像匹配和图像融合的方法对检测结果图像和常规图像进行处理，确定待显示图像。
在再一个可选的实现方式中，上述“处理器”还用于，在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。
在再一个可选的实现方式中，上述终端还包括：显示屏，用于显示待显示图像。
在再一个可选的实现方式中，上述“处理器”还可以用于：在皮肤检测模式下确定待检测特征；在皮肤检测模式下根据待检测特征确定拍摄待检测原图像的摄像头，摄像头包括：彩色摄像头和黑白摄像头中的至少一个。
根据不同的待检测特征，提供了不同的预置规则进行处理，例如：根据肤色对应的预设规则处理之后，可以减少皮肤噪点；根据色斑和红区中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再经过RBX空间的B通道(brown通 道)凸显色斑，R(red通道)通道可以凸显红区，根据痤疮对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再通过HSV空间图像凸显痤疮特征；根据油分对应的预置规则处理，可以通过皮肤反光部分完成油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分；根据眼部特征和毛孔中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度突出特征区域。
在再一个可选的实现方式中，若待检测特征包括：肤色、色斑、红区和痤疮中的至少一种，则处理器在皮肤检测模式下确定拍摄待检测原图像的摄像头为彩色摄像头；若待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种，则处理器在皮肤检测模式下确定拍摄待检测原图像的摄像头为黑白摄像头。
在再一个可选的实现方式中，上述“处理器”包括：图像信号处理器和中央处理器，该图像信号处理器，用于在皮肤检测模式下根据待检测特征确定待检测原图像；根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；确定常规图像。
中央处理器，用于对检测专用图像进行检测，确定检测结果图像；根据检测结果图像和常规图像，确定待显示图像。
第四方面，本申请实施例提供了一种计算机可读存储介质，包括计算机可读指令，当计算机读取并执行所述计算机可读指令时，使得计算机执行前述第一方面以及其可选地实现中的方法。
第五方面，本申请实施例提供了一种计算机程序产品，包括计算机可读指令，当计算机读取并执行所述计算机可读指令，使得计算机执行前述第一方面以及其可选地实现中的方法。
附图说明
下面将参照所示附图对本申请实施例进行更详细的描述。
图1为图像信号处理器处理图像的方法流程图；
图2为本发明实施例提供的一种图像检测终端结构示意图；
图3为本发明实施例提供的一种图像检测的方法流程图；
图4(a)为本发明实施例提供的一种经处理RAW格式的待检测原图像；
图4(b)为本发明实施例提供的一种JPEG格式的图像；
图5(a)为本发明实施例提供的一种模式选择的界面图；
图5(b)为本发明实施例提供的另一种模式选择的界面图；
图6(a)为本发明实施例提供的一种含有痤疮的待检测原图像的示意图；
图6(b)为本发明实施例提供的一种含有痤疮的检测专用图像的示意图；
图7(a)为本发明实施例提供的一种油分含量大的图像的示意图；
图7(b)为本发明实施例提供的一种油分含量正常的图像的示意图；
图8为本发明实施例提供的一种痤疮检测的检测结果图像的示意图；
图9为本发明实施例提供的一种含有色斑的待显示图像的示意图；
图10为本发明实施例提供的一种含有眼部特征的待显示图像的示意图；
图11为本发明实施例提供的一种含有毛孔的待显示图像的示意图；
图12为本发明实施例提供的一种含有待显示图像的显示界面图；
图13为本发明实施例提供的另一种含有待显示图像的显示界面图；
图14为本发明实施例提供的一种图像检测的方法流程图；
图15为本发明实施例提供的另一种图像检测的方法流程图；
图16为本发明实施例提供的又一种图像检测的方法流程图；
图17为本发明实施例提供的一种图像检测的装置示意图；
图18为本发明实施例提供的一种终端示意图。
具体实施方式
为便于对本发明实施例的理解，下面将结合附图以具体实施例做进一步的解释说明，实施例并不构成对本发明实施例的限定。
本发明实施例提供一种图像检测的方法、装置及终端，通过在皮肤检测模式下对待检测原文件进行处理和检测，得到检测结果图像，再将该检测结果图像和常规图像进行匹配和融合得到了待显示图像。解决了成像质量不满足健康检测的检测精度要求的问题，所以本方案提高了成像的质量并使检测结果更加准确。
图1为图像信号处理器处理图像的方法流程图。如图1所示，图像信号处理器(internet service provider，ISP)对格式为RAW的未经加工图像进行如下处理。其中，该图像信号处理器可以设置在终端中。
S101：ISP接收格式为RAW的未经加工图像即下文中出现的待检测原图像，并进行预处理，处理过程如下：对格式为RAW的待检测原图像进行噪点调整(hot pixel correction)、去马赛克(demoscal)、降噪(noise reduction)、图像斑点调整(shading correction)、图像加工(geometric correction)、色彩修改(color correction)、色调曲线调整(tone curve adjustment)以及勾边轮廓增强(edge enhancement)，得到至少一个YUV格式的图像输出。
S102：ISP将预处理后的图像发送至JPEG编码器，该JPEG编码器对预处理后的图像进行压缩，并输出JPEG格式的图像。现有技术中，利用该处理未经加工图像的方法得到的JPEG图像无法满足后续皮肤特征检测的精度要求，具体表现在：第一，在S102中，格式为RAW的未经加工图像进行有损压缩得到JPEG格式的图像，一些皮肤细节特征经JPEG编码器压缩后变得不明显，使后续检测JPEG格式的图像的难度加大，导致结果不准确。第二，经过S101和S102处理过后，其中的JPEG图像每个像素只保留了8位(bit)颜色信息，而RAW格式的颜色信息为10-16位(bit)，由此可知，导致了JPEG图像的色阶大大减少，后期对对比度、光线等的可调整范围也会相应变小，导致结果不准确。第三、此方法处理之后得到的JPEG图像的白平衡调整范围有限，且色温调整通常会损失细节信息，导致之后的结果不准确。
S103：ISP也可以直接将该格式为RAW的未经加工图像直接输出。
S104：3A算法。具体包括：3A技术即自动对焦(AF)、自动曝光(AE)和自动白平衡(AWB)。3A数字成像技术利用了AF自动对焦算法、AE自动曝光算法及AWB自动白平衡算法来实现RAW图像对比度最大、改善主体拍摄物过曝光或曝光不足情况、使画面在不同光线照射下的色差得到补偿，从而呈现较高画质的图像信息。采用了3A数字成像技术的摄像机能够很好的保障图像精准的色彩还原度，呈现完美的日 夜监控效果。
图2为本发明实施例提供的一种图像检测终端结构示意图。如图2所示，该终端可以包括手机、平板电脑、笔记本电脑、个人数字助理(personal digital assistant，PDA)、销售终端(point of sales，POS)以及车载电脑等。该终端至少可以包括摄像头210、处理装置、图像信号处理器230、显示器240和通信总线250，摄像头210、处理装置、图像信号处理器(image signal processor，ISP)230、显示器240通过通信总线250相互连接，并完成相互之间的通信。
该摄像头210用于拍摄图像，该图像包括：常规图像和未经处理的RAW格式的图像(即下文中的待检测原图像)。需要说明的是，该摄像头可以为单一摄像头，也可以是双摄像头，该双摄像头可以包括：彩色摄像头和黑白摄像头。
该处理装置可以包括：接收器221、处理器222和存储器223。该接收器221用于接收常规图像和检测专用图像。该存储器223用户存储检测专用图像。
该图像信号处理器230，用于调用操作指令和接收未经处理的RAW格式的图像，即摄像头210拍摄的原图像并执行以下操作：
在皮肤检测模式下根据待检测特征确定待检测原图像；
根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；
确定常规图像；
对检测专用图像进行检测，确定检测结果图像；
根据检测结果图像和常规图像，确定待显示图像。
需要说明的是，上述方法可以是在图像信号处理器230中进行处理。也可以是在图像信号处理器230和处理装置中分别进行处理，例如上述“在皮肤检测模式下根据待检测特征确定待检测原图像；根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；以及确定常规图像”的步骤在图像信号处理器230中进行处理，然后“对检测专用图像进行检测，确定检测结果图像；根据检测结果图像和常规图像，确定待检测图像”的步骤在处理装置中进行处理。
需要说明的是，上述步骤可以理解为在图像信号处理器230中新增的并行方法，也就说，获取检测专用图像的步骤并没有采取除图像信号处理器230以外的处理器进行处理，该过程可以是在图像信号处理器230完成的。一般来说如果是利用除图像信号处理器230以外的处理器处理，是在拍摄并保存之后，该处理器将调用终端存储器233中保存的常规图像以及检测专用图像，然后该处理器再对检测专用图像进行处理生成检测结果图像。但从本申请保护技术的技术方案来看，在图像信号处理器230中实现上述操作为优选方案，理由如下，在图像信号处理器230中实现上述操作，不需要额外存储RAW图像，占用空间小。其次，在图像信号处理器230中实现，成本低且处理速度快，提高用户体验。
在上述“确定常规图像”的步骤中，该步骤获得的常规图像是与生成检测专用图像同时生成的。
在本发明实施例中，该图像信号处理器230，还可以用于根据待检测原图像和待检测特征确定检测专用图像，该确定的过程可以包括：对该待检测原图像进行光照均衡、白平衡调整、图像平滑、对比度变换、图像锐化或颜色空间变换中的至少一个处理确定输出 的检测专用图像。
该装置中的处理装置还可以包括：处理器和存储器。该处理器可以是中央处理单元(central processing unit，CPU)，该处理器还可以是其他通用处理器、数字信号处理器(digital signal processor，DSP)、专用集成电路(application specific integrated circuit，ASIC)、现成可编程门阵列(field-programmable gate array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。该处理器还可以接收图像信号处理器230确定的检测专用图像以及常规图像，对该检测专用图像检测并确定检测结果图像，该处理器匹配和融合检测结果图像和常规图像，确定待显示图像。
该存储器可以包括只读存储器和随机存取存储器，具体地，可以存储检测专用图像，并向图像信号处理器230提供指令和数据。存储器的一部分还可以包括非易失性随机存取存储器。例如，存储器还可以存储设备类型的信息。
通信总线250除包括数据总线之外，还可以包括电源总线、控制总线和状态信号总线等。但是为了清楚说明起见，在图中将各种总线都标为通信总线250。
此外，该显示器240用于显示待显示图像。尽管未示出，该终端还可以包括蓝牙、天线、麦克风等，在此不再赘述。本领域技术人员可以理解，图2中示出的终端结构并不构成对终端的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。
为了对下述方法实施例进行更详细的描述，该方法的执行主体可以为图2中所示的终端，为了方便描述，下述实施例以手机为例，实施例结合图3至图16进行详细描述。
图3为本发明实施例提供的一种图像检测的方法流程图。如图3所示，该方法300可以对全身的皮肤进行检测，为了更清楚的介绍，下述方法以面部皮肤为例，该方法300可以具体包括以下步骤：
S310：终端在皮肤检测模式下根据待检测特征确定待检测原图像。
其中，该待检测原图像为RAW格式的图像(即上述未经处理的RAW格式的图像)，如图4(a)所示，401的位置显示出的皮肤状况更为清晰，而图4(b)为现有技术中对JPEG的图像进行检测，其中该JPEG的图像显示的402部分的细节已经模糊，如果按照现有技术得到的图像进行检测，将无法得到更为准确的检测结果。
具体地，本发明实施例提供的皮肤检测模式可以是预设的模式。打开皮肤检测模式进行拍摄待检测原图像；也可以通过非皮肤检测模式(即普通拍摄模式)进行拍摄。该待检测特征包括：肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种。该皮肤检测预设的模式可以通过3A算法进行处理，其3A算法包括：自动曝光(AE)、自动白平衡(AWB)、自动对焦(AF)，实现一个负反馈闭环控制系统，即状态硬件模块输出光度值、光圈值或灰度空间，然后对应的3A算法进行反馈到相应的处理单元做处理，直到相应的值稳定下来，再进行拍摄待检测原图像。
其中，在一个可实现的实施例中，若待检测特征包括：肤色、色斑、红区和痤疮中的至少一种，则终端在皮肤检测模式下采用彩色摄像机拍摄检测原图像，然后该终端确定检测原图像中的待检测原图像，该待检测原图像为彩色待检测原图像。
若所述待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种，则终端在皮肤 检测模式下采用黑白摄像头拍摄检测原图像，然后该终端确定检测原图像中的待检测原图像，该待检测原图像为黑白待检测原图像。
例如，如图5(a)所示，用户在所述终端上选择皮肤检测模式，其中，图5(a)中的模式选择可以包括多种模式，如普通模式、皮肤检测模式和其他模式，该皮肤检测模式可以包括：第一皮肤检测模式、第二皮肤检测模式等多种皮肤检测模式。例如：第一皮肤检测模式可以包括的待检测特征为：肤色、色斑、红区和痤疮中的至少一种；第二皮肤检测模式可以包括的待检测特征为：油分、皱纹、眼部特征和毛孔中的至少一种。如图5(b)该终端在皮肤检测模式下显示检测特征选项，为用户提供待检测特征选项，该终端根据用户的选择确定待检测特征。如果用户没有进行选择检测特征，则终端自动默认对全部待检测特征进行检测。该终端摄像头可以为双摄像头，该双摄像头可以包括：彩色摄像头和黑白摄像头。
待检测特征可以有多种组合方式，分别对应不同的皮肤检测模式，例如：肤色、色斑、红区和痤疮的组合为第一组合方式，油分、皱纹、眼部特征和毛孔的组合为第二组合方式，根据待检测特征的组合方式选择适合的摄像头，根据选择后的摄像头拍摄更适合待检测特征的待检测原图像(即未经处理的RAW格式的图像)。该摄像头可以是在皮肤检测模式下调用的，用户可以自由选择拍摄模式，拍摄模式对应的待检测特征也可以用户选择，在此不做限定。当用户不对待检测特征进行选择时，该终端自动默认为对待检测特征进行全部检测。该终端确定的摄像头也可以为双摄像头，该双摄像头可以包括：彩色摄像头和黑白摄像头。
在另一个可实现的实施例中，若待检测特征包括：肤色、色斑、红区和痤疮中的至少一种，则终端在皮肤检测模式下采用双摄像头进行拍摄待检测原图像(即彩色摄像头和黑白摄像头同时进行拍摄)，然后该终端仅提取彩色摄像头对应的未经处理的RAW格式的图像作为待检测原图像。
若待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种，则终端在皮肤检测模式下采用双摄像头进行拍摄待检测原图像(即彩色摄像头和黑白摄像头同时进行拍摄)，然后该终端仅提取黑白摄像头对应的未经处理的RAW格式的图像作为待检测原图像。
S320：终端根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像。
具体地，该检测专用图像的格式可以为JPEG格式。需要说明的是，该步骤是在终端获取到待检测原图像之后直接根据待检测特征对应的预置规则处理之后生成的检测专用图像，可以理解为是经过拍摄流程之后直接生成的检测专用图像。而不是先获得待检测原图像，对待检测原图像进行存储，然后采用待检测特征对应的预置规则进行后期处理来生成检测专用图像的。
该预置规则包括多个子规则。可以理解的是，该终端根据待检测特征确定该待检测特征对应的预置规则，经过该检测特征对应的预置规则处理之后得到该待检测特征对应的处理之后的检测专用图像，该预置规则具体包括如下：
该肤色对应的预置规则，包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。例如：根据肤色对应的预设规则处理之后，可以减少皮肤噪点。
该色斑和红区中的至少一种对应的预置规则，包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则。例如：根据色斑和红区中的至少一种对应的预置规则，可以通过增大对比度和提高图像锐化程度至预设值，再经过RBX空间的B通道(brown通道)凸显色斑，R(red通道)通道可以凸显红区。
该痤疮对应的预置规则，包括：对比度变换、图像锐化和HSV空间图像中的至少一个子规则。例如：图6(a)所示的为待检测原图像，通过痤疮对应的预置规则处理后，得到的检测专用图像可以为图6(b)所示。经过处理后的痤疮的位置、大小等信息更为突出。例如：根据痤疮对应的预置规则，可以通过增大对比度和提高图像锐化程度至预设值，再通过HSV空间图像凸显痤疮特征。
该油分对应的预置规则，包括：对比度变换。例如：为了明显区分油分大的皮肤和正常皮肤，本实施例提供了如图7所示的图像，图7(a)为油分大的皮肤，有光亮的地方为油份大的。例如：根据油分对应的预置规则，可以通过皮肤反光部分完场油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分。
该眼部特征和毛孔中的至少一种对应的预置规则，包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。例如：根据眼部特征和毛孔中的至少一种对应的预置规则，可以通过增大对比度和提高图像锐化程度突出特征区域。
其中，该白平衡调整的调整幅度为预设值，RBX颜色空间变换和HSV空间图像的变换空间为预设值。上述的检测专用图像可以为多个图像，该检测专用图像的图像格式可以为JPEG格式。例如：该预设值需要根据大量实验数据提前设定。
S330：该终端确定常规图像。
具体地，该常规图像可以通过常规预处理直接获得，也可以通过融合常规图像的至少两个子图像获得。其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种，该常规图像的格式可以为JPEG格式，该子图像的格式可以为JPEG格式。
上述融合常规图像的多个子图像的步骤可以包括：对至少两个子图像分别进行像素级别的同步，然后对齐，最后用图像融合算法提取至少两个图像数据中的优势部分，并合成一张图像，其中优势部分可以包括：彩色常规子图像中的颜色部分和黑白常规子图像中的阴影的细节部分。
S340：终端对检测专用图像进行检测，确定检测结果图像。
具体地，终端利用皮肤检测算法对至少一个检测专用图像进行检测，确定至少一个检测结果图像。其中，多个待检测特征中的任一个对应一个检测专用图像，一个检测专用图像对应一个皮肤检测算法进行检测，确定一个检测结果图像。该检测结果图像可以包括：待检测特征的标记。该检测专用图像可以为多个，该皮肤检测算法可以包括多个子算法，该检测结果图像可以为多个图像，该检测结果图像的图像格式可以包括JPEG图像。该皮肤检测算法可以包括：
若该肤色对应的检测专用图像为肤色检测专用图像，则利用肤色对应的肤色检测算法对肤色检测专用图像进行检测，该肤色检测算法可以包括：根据预设的标准图片对获取到的肤色检测专用图像进行颜色校准，然后将该肤色检测专用图像与标准肤色色号作比较，确定最接近的肤色，确定含有肤色特征标记的检测结果图像。
若该色斑和红区中的至少一种对应的检测专用图像为色斑和红区中的至少一种检测专用图像，则利用色斑和红区中的至少一种对应的色斑和红区中的至少一种检测算法对色斑和红区中的至少一种检测专用图像进行检测，该色斑和红区中的至少一种检测算法可以包括：对色斑和红区中的至少一种检测专用图像，提取上述的RBX图像中的R通道(或B通道)分量，对该分量做图像处理(如边缘检测)，确定含有该色斑和红区中的至少一种标记的检测结果图像。
若该痤疮对应的检测专用图像为痤疮检测专用图像，则利用痤疮对应的痤疮检测算法对痤疮检测专用图像进行检测，该痤疮检测算法可以包括：对痤疮检测专用图像，即上述的HSV空间图像做图像处理，确定含有痤疮标记的检测结果图像。如图8所示，图8为痤疮检测的检测结果图像，该检测结果图像可以包括：待检测特征的标记(图8中示出的矩形框为标记，标记的形状可以是除矩形外的多种形状，在此不再限定)。
若该油分对应的检测专用图像为油分检测专用图像，则利用油分对应的油分检测算法对油分检测专用图像进行检测，该油分检测算法可以包括：通过图像处理手段检测油分检测专用图像中亮度较大的区域，计算该区域占面部区域的面积从而确定油份大小，确定含有油分特征标记的检测结果图像。
若该眼部特征和毛孔中的至少一种对应的检测专用图像为眼部特征和毛孔中的至少一种检测专用图像，则利用眼部特征和毛孔中的至少一种对应的眼部特征和毛孔中的至少一种检测算法对眼部特征和毛孔中的至少一种检测专用图像进行检测，确定含有眼部特征和毛孔中的至少一种特征标记的检测结果图像，该眼部特征和毛孔中的至少一种检测算法可以包括：图像处理类的算法，如边缘检测、图像滤波等，也可以是机器学习类的方法。
S350：该终端根据该检测结果图像和常规图像，确定待显示图像。
具体地，该终端利用图像匹配和图像融合的方法对该至少一个检测结果图像和常规图像进行处理，获得该待显示图像。
其中，图像匹配是通过匹配算法在至少一个检测结果图像和常规图像之间识别同名点。例如：在图像匹配中通过比较两幅或多幅图像中的目标区域和搜索区中相同大小的窗口的相关系数，取搜索区中相关系数最大所对应的窗口中心点作为同名点。图像融合是对至少一个检测结果图像和常规图像分别进行像素级别的同步，然后对齐，最后用图像融合算法提取至少两个图像数据中的优势部分，并合成一张图像，最大限度的提取各自信道中的有利信息，最后综合成高质量的图像，以提高图像信息的利用率、提升原始图像的空间分辨率和光谱分辨率，利于检测。上述图像融合过程与S330步骤中的融合类似，只是两者融合的原图片不同，采用的方法都是利用图像融合算法进行融合。
需要说明的是，检测结果图像可以是包括对待检测特征进行标记的图像；即可以理解为是将检测专用图像含有待检测特征部分的图像提取出来而形成的图像，并对该待检测特征部分进行标记，具体如图8中的矩形框所示。
待显示图像可以是将含有对待检测特征进行标记的图像和常规图像进行匹配融合后的图像，具体如图9-图11所示。将常规图像中与检测结果图像对应的区域与检测结果图像进行匹配，并使用该检测结果图像覆盖/替换该常规图像中与检测结果图像对应的区域，从而得到待显示区域。
例如：该待显示图像可以为如图9所示。图9可以为色斑对应的待显示图像，901可以显示为色斑的位置和大小，经过色斑对应的预置规则和检测算法，可以使色斑的位置显示的更加明显。图10可以为眼部特征的待显示图像，1001可以显示为眼部特征的位置和大小，经过眼部特征对应的预置规则和检测算法，可以突出眼部特征存在的问题。图11可以为毛孔的待显示图像，1101可以显示为毛孔的位置和大小，经过毛孔对应的预置规则和检测算法，可以突出毛孔的位置以便于用户观看。此外，待显示图像的图像格式可以为JPEG图像。
S360：该终端显示待显示图像。
具体地，该终端显示的内容可以包括待显示图像，也可以包括待显示图像和根据待显示图像得到的文字描述和护理信息的至少一种。
其中，该护理信息的形式是多种多样的，例如：该护理信息的形式可以是查找预设的专家建议库，然后得到与检测结果图像所对应的护理建议；也可以是根据检测结果图像推送相对应的链接等(例如：专家博客或网站)；也可以选择线上咨询人员根据待检测结果图像提供面对面的线上护理建议，对此在本发明实施例中不作任何限制。
除了上述的文字描述和护理信息的至少一种，该终端显示的内容还可以包括：根据检测结果图像得到的检测结果在同龄人中的排名或评分等信息，需要说明的是，该排名或评分都是根据检测结果的基础上派生出的后续评价，对此在本发明实施例中不作任何限制。
例如：如图12所示的显示界面，该显示界面可以只包括待显示图像1201，该待显示图像可以包括：检测结果图像和护理建议等信息。也可以是如图13所示的待显示图像，该待显示图像包括至少两个部分：检测结果图像1301和建议1302，该建议1302可以包括：检测结果、护理建议、在同龄人的皮肤健康状态的排名以及评分等信息。在本实施例中，不对显示图像作任何限定，也不对建议部分做任何限定，该建议部分包括1201部分和1302的建议部分。
为了进一步理解本发明实施例提供的图像检测的方法，结合图13-图15，对本发明实施例提供的图像检测的方法进行详细说明，具体如下。
图14为本发明实施例提供的一种图像检测的方法流程图。如图14所示，若用户选择肤色、色斑、红区和痤疮中的至少一种，则采用如图14所示的方法对图像进行检测。具体地，例如：当用户打开皮肤健康检测应用后，终端可以启动相机拍摄功能，在该皮肤健康检测应用提供的皮肤检测模式下进行拍摄。也可以直接启动终端的相机拍摄功能，然后选择皮肤检测模式进行拍摄。
用户可以选择在该皮肤健康检测模式中的待检测特征，例如：肤色。终端确定待检测特征为肤色，则终端调用摄像头进行拍摄，该终端获取全面的拍摄图像的的原始数据，其中，该摄像头包括：彩色摄像头和/或黑白摄像头。同时，该终端读取原始数据中的RAW彩色待检测原图像1401，并采用肤色对应的预置规则1对该彩色待检测原图像1401进行处理，得到检测专用图像1402。该肤色对应的预置规则1可以包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。例如：根据肤色对应的预设规则进行处理之后，可以减少皮肤噪点。
该终端采用该肤色对应的检测算法1对待检测专用图像1402进行处理，得到检测结 果图像1403。其中，该肤色对应的检测算法1可以包括：根据预设的标准图片对获取到的肤色检测专用图像进行颜色校准，然后将该肤色检测专用图像与标准肤色色号作比较，确定最接近的肤色，将比较出的最接近的肤色作为待检测专用图像1402的肤色标记，得到该检测结果图像1403，该检测结果图像1403包括：肤色的标记信息，该肤色的标记信息可以包括：肤色的色调等。
该终端将常规图像子图像1404、和常规图像子图像1504与该检测结果图像1403进行匹配和融合。获得待显示图像1406。其中，该常规图像子图像1404是终端利用通用预处理对彩色待检测原图像1401进行处理得到的，该彩色待检测原图像1401是可以由彩色摄像头拍摄。该常规图像子图像1504是终端利用通用预处理对黑白检测原图像1501进行处理得到的，该黑白待检测原图像1501是可以由黑白摄像头拍摄。
该终端显示待显示图像1406。
上述过程中，该待显示图像1406可以包括：该终端显示的内容可以包括待显示图像1406，也可以包括待显示图像1406和根据待显示图像1406得到的文字描述和护理信息的至少一种。其中，该护理信息的形式是多种多样的，例如：该护理信息的形式可以是查找预设的专家建议库，然后得到与检测结果图像所对应的护理建议；也可以是根据检测结果图像推送相对应的链接等(例如：专家博客或网站)；也可以选择线上咨询人员根据待检测结果图像提供面对面的线上护理建议，对此在本发明实施例中不作任何限制。
除了上述的文字描述和护理信息的至少一种，该终端显示的内容还可以包括：根据检测结果图像得到的检测结果在同龄人中的排名或评分等信息，需要说明的是，该排名或评分都是根据检测结果的基础上派生出的后续评价。
图15为本发明实施例提供的另一种图像检测的方法流程图。如图15所述，若用户选择油分、皱纹、眼部特征和毛孔中的至少一种，则采用如图15所示的方法对图像进行检测。
具体地，例如：当用户打开皮肤健康检测应用后，终端可以启动相机拍摄功能，在该皮肤健康检测应用提供的皮肤检测模式下进行拍摄。也可以直接启动终端的相机拍摄功能，然后选择皮肤检测模式进行拍摄。
用户可以选择在该皮肤健康检测模式中的待检测特征，例如：油分。终端确定待检测特征为油分，则终端调用摄像头进行拍摄，该终端获取全面的拍摄图像的的原始数据，其中，该摄像头包括：彩色摄像头和/或黑白摄像头。同时，该终端读取原始数据中的RAW黑白待检测原图像1501，并采用油分对应的预置规则2对该彩色待检测原图像1501进行处理，得到检测专用图像1502。该油分对应的预置规则2可以包括：对比度变换。例如：为了明显区分油分大的皮肤和正常皮肤，根据油分对应的预置规2处理之后，可以通过皮肤反光部分完场油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分。
该终端采用该油分对应的检测算法2对待检测专用图像1502进行处理，得到检测结果图像1503。其中，该油分对应的检测算法2可以包括：通过图像处理手段检测油分检测专用图像中亮度较大的区域，计算该区域占面部区域的面积从而确定油份大小，将计算出的油分大小作为待检测专用图像1502的油分标记，得到该检测结果图像1503，该检测结 果图像1503包括：油分的标记信息，该油分的标记信息可以包括：油分在整个面部的位置和含油量等。
该终端将常规图像子图像1504、常规图像子图像1404与该检测结果图像1503进行匹配和融合。获得待显示图像1506。其中，该常规图像子图像1504是终端利用通用预处理对黑白待检测原图像1501进行处理得到的，该黑白待检测原图像1501可以由黑白摄像头拍摄。该常规图像子图像1404是终端利用通用预处理对彩色检测原图像1401进行处理得到的，该彩色待检测原图像1401可以由彩色摄像头拍摄。
该终端显示待显示图像1506。
上述过程中，该待显示图像1506可以包括：该终端显示的内容可以包括待显示图像1506，也可以包括待显示图像1506和根据待显示图像1506得到的文字描述和护理信息的至少一种。其中，该护理信息的形式是多种多样的，例如：该护理信息的形式可以是查找预设的专家建议库，然后得到与检测结果图像所对应的护理建议；也可以是根据检测结果图像推送相对应的链接等(例如：专家博客或网站)；也可以选择线上咨询人员根据待检测结果图像提供面对面的线上护理建议，对此在本发明实施例中不作任何限制。
除了上述的文字描述和护理信息的至少一种，该终端显示的内容还可以包括：根据检测结果图像得到的检测结果在同龄人中的排名或评分等信息，需要说明的是，该排名或评分都是根据检测结果的基础上派生出的后续评价。
图16为本发明实施例提供的又一种图像检测的方法流程图。如图16所示，若用户选择肤色、色斑、红区和痤疮中的至少一种和油分、皱纹、眼部特征和毛孔中的至少一种，则采用如图16所示的方法对图像进行检测。
例如：当用户打开皮肤健康检测应用后，终端可以启动相机拍摄功能，在该皮肤健康检测应用提供的皮肤检测模式下进行拍摄。也可以直接启动终端的相机拍摄功能，然后选择皮肤检测模式进行拍摄。用户可以选择在该皮肤健康检测模式中的待检测模式，例如：肤色和油分。终端在皮肤检测模式下选择肤色和油分两个待检测特征，选择待检测特征之后，该终端调用彩色摄像头拍摄彩色待检测原图像1401以及黑白摄像头拍摄黑白待检测原图像1501。该彩色待检测原图像1401和黑白待检测原图像1501的图像格式可以为RAW格式，该过程，是在格式为RAW图像上进行处理，通过该过程得到的结果要比在压缩后的JPEG格式的图像处理后的结果更加准确。其中，该皮肤检测模式可以是开启皮肤健康检测应用之后，启动相机拍摄功能。也可以是直接调用相机应用，选择皮肤检测模式进行拍摄。该终端根据肤色对应的预置规则1对彩色待检测原图像进行处理，得到包含肤色的检测专用图像1402。该肤色对应的预置规则1可以包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。例如：根据肤色对应的预设规则处理之后，可以减少皮肤噪点。该终端利用肤色对应的检测算法1对包含肤色的检测专用图像1402进行检测，得到包含肤色的检测结果图像1403。该肤色对应的检测算法1可以包括：根据预设的标准图片对获取到的肤色检测专用图像1402进行颜色校准，然后将该肤色检测专用图像1402与标准肤色色号作比较，确定最接近的肤色。
终端利用常规预处理的方法对彩色待检测原图像1401进行处理，得到常规图像子图像1404，该常规图像子图像1404的图像格式可以为JPEG格式。
该终端根据油分对应的预置规则2对黑白待检测原图像1501进行处理，得到包含油分的检测专用图像1502。该油分对应的预置规则2可以包括：对比度变换。例如：根据油分对应的预置规则2处理，可以通过皮肤反光部分完场油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分。该终端利用油分对应的检测算法2对包含油分的检测专用图像1502进行检测，得到包含油分的检测结果图像1503。该油分对应的检测算法2可以包括：通过图像处理手段检测油分检测专用图像1502中亮度较大的区域，计算该区域占面部区域的面积从而确定油份大小。
终端利用常规预处理的方法对黑白待检测图像1501进行处理，得到常规图像子图像1504，该常规图像子图像1504的图像格式可以为JPEG格式。
该终端将该常规图像子图像1404和常规图像子图像1504进行融合，得到常规图像。该终端将上述包含肤色的检测结果图像1401、包含油分的检测结果图像1501以及常规图像1605进行匹配和融合，获得待显示图像1606。
该终端显示待显示图像。具体内容详见S360，在此不再赘述。需要进一步说明的是，显示的图像可以在存储器调用观看，也可以在存储器调用之后再次处理。图17为本发明实施例提供的一种图像检测的装置示意图。
如图17所示，本发明实施例提供了一种图像检测的装置。该装置包括：
第一确定单元1701，用于在皮肤检测模式下根据待检测特征确定待检测原图像。
第二确定单元1702，用于根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像。所述第二确定单元1702还用于，确定常规图像。
检测单元1703，用于对检测专用图像进行检测，确定检测结果图像。
匹配单元1704，用于根据检测结果图像和常规图像，确定待显示图像。
上述待检测特征包括：肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种，其中，该该眼部特征可以包括：眼袋、黑眼圈和眼部细纹中的至少一个。
第一确定单元1701还可以用于，该待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
可选地，当待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，待检测原图像为彩色待检测原图像；当待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种时，待检测原图像为黑白待检测原图像。
由于该待检测特征的种类不同，所以在可选地实现方式中可以针对该待检测特征种类的特性提供不同的待检测原图像，为后期处理提供一个精度较高的未经处理过的原图像，保证待检测原图像的高精度。
第二确定单元1702可以用于，融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。在再一个可选的实现方式中，肤色对应的预置规则，包括：
当待检测特征为肤色时，该肤色对应的预置规则包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。
当待检测特征为色斑和红区中的至少一种时，色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则。
当待检测特征为痤疮时，痤疮对应的预置规则包括：对比度变换、图像锐化和HSV空 间图像中的至少一个子规则。
当待检测特征为油分时，油分对应的预置规则包括：对比度变换。
当待检测特征为眼部特征和毛孔中的至少一种时，眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
其中，白平衡调整的调整幅度为预设值，RBX颜色空间变换和HSV空间图像的变换空间为预设值。
匹配单元1704还用于，利用图像匹配和图像融合的方法对检测结果图像和常规图像进行处理，确定待显示图像。
第一确定模块1701还用于，在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。
上述装置还包括：显示模块，用于显示待显示图像。
第一确定单元1701还可以用于：在皮肤检测模式下根据待检测特征确定待检测特征；在皮肤检测模式下根据待检测特征确定拍摄待检测原图像的摄像头，摄像头包括：彩色摄像头和黑白摄像头中的至少一个。
根据不同的待检测特征，提供了不同的预置规则进行处理，例如：根据肤色对应的预设规则处理之后，可以减少皮肤噪点；根据色斑和红区中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再经过RBX空间的B通道(brown通道)凸显色斑，R(red通道)通道可以凸显红区，根据痤疮对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再通过HSV空间图像凸显痤疮特征；根据油分对应的预置规则处理，可以通过皮肤反光部分完成油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分；根据眼部特征和毛孔中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度突出特征区域。
若待检测特征包括：肤色、色斑、红区和痤疮中的至少一种，则在皮肤检测模式下确定拍摄待检测原图像的摄像头为彩色摄像头；若待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种，则在皮肤检测模式下确定拍摄待检测原图像的摄像头为黑白摄像头。
图18为本发明实施例提供的一种终端示意图。如图18所示，本发明实施例提供了一种终端。该终端包括：处理器1801，用于在皮肤检测模式下根据待检测特征确定待检测原图像；处理器还用于，根据待检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；以及确定常规图像；处理器还用于，对检测专用图像进行检测，确定检测结果图像；处理器还用于，根据检测结果图像和常规图像，确定待显示图像。
本方案中，通过处理器在皮肤检测模式下对待检测原文件进行处理和检测，得到检测专用图像，处理器还可以再根据检测专用图像确定检测结果图像，再对该检测结果图像和常规图像进行处理得到了待显示图像。解决了成像质量不满足健康检测的检测精度的要求，以及不能为用户随时随地提供便捷、准确和专业的皮肤健康检测和评估的问题。
待检测特征包括：肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种，其中，该该眼部特征可以包括：眼袋、黑眼圈和眼部细纹中的至少一个。
上述待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
具体地，当待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，待检测原图像为彩色待检测原图像；当待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种 时，待检测原图像为黑白待检测原图像。
由于该待检测特征的种类不同，所以在可选的实现方式中可以针对该待检测特征种类的特性提供不同的待检测原图像，为后期处理提供一个精度较高的未经处理过的原图像，保证待检测原图像的高精度。
处理器还可以用于，融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。具体地，当待检测特征为肤色时，该肤色对应的预置规则包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则。
当待检测特征为色斑和红区中的至少一种时，该色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则。
当待检测特征为痤疮时，该痤疮对应的预置规则包括：对比度变换、图像锐化和HSV空间图像中的至少一个子规则。
当待检测特征为油分时，该油分对应的预置规则包括：对比度变换。
当待检测特征为眼部特征和毛孔中的至少一种时，该眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
其中，白平衡调整的调整幅度为预设值，RBX颜色空间变换和HSV空间图像的变换空间为预设值。
处理器可以用于，利用图像匹配和图像融合的方法对检测结果图像和常规图像进行处理，确定待显示图像。
处理器可以用于，在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。
上述终端还包括：显示屏，用于显示待显示图像。
此外，处理器还可以用于：在皮肤检测模式下确定待检测特征；在皮肤检测模式下根据待检测特征确定拍摄待检测原图像的摄像头，摄像头包括：彩色摄像头和黑白摄像头中的至少一个。
根据不同的待检测特征，提供了不同的预置规则进行处理，例如：根据肤色对应的预设规则处理之后，可以减少皮肤噪点；根据色斑和红区中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再经过RBX空间的B通道(brown通道)凸显色斑，R(red通道)通道可以凸显红区，根据痤疮对应的预置规则处理，可以通过增大对比度和提高图像锐化程度至预设值，再通过HSV空间图像凸显痤疮特征；根据油分对应的预置规则处理，可以通过皮肤反光部分完成油分检测，通过增大对比度至预设值，可以进一步凸显反光部分，根据反光部分检测油分；根据眼部特征和毛孔中的至少一种对应的预置规则处理，可以通过增大对比度和提高图像锐化程度突出特征区域。
具体地，若待检测特征包括：肤色、色斑、红区和痤疮中的至少一种，则处理器在皮肤检测模式下确定拍摄待检测原图像的摄像头为彩色摄像头；若待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种，则处理器在皮肤检测模式下确定拍摄待检测原图像的摄像头为黑白摄像头。
在另外一种可以实现的方式中，上述处理器可以包括：图像信号处理器和中央处理器。
图像信号处理器，用于在皮肤检测模式下根据待检测特征确定待检测原图像；根据待 检测特征对应的预置规则，对待检测原图像进行处理得到检测专用图像；确定常规图像。
中央处理器，用于对检测专用图像进行检测，确定检测结果图像；根据检测结果图像和常规图像，确定待显示图像。
需要说明得是，本方案的优选方案包括：上述处理器优选图像信号处理器。在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时，全部或部分地产生按照本发明实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质，(例如，软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质(例如固态硬盘Solid State Disk(SSD))等。
以上所述的具体实施方式，对本申请的目的、技术方案和有益效果进行了进一步详细说明，所应理解的是，以上所述仅为本申请的具体实施方式而已，并不用于限定本申请的保护范围，凡在本申请的技术方案的基础之上，所做的任何修改、等同替换、改进等，均应包括在本申请的保护范围之内。

Claims (34)

  1. 一种图像处理的方法，其特征在于，包括：
    终端在皮肤检测模式下根据待检测特征确定待检测原图像；
    所述终端根据所述待检测特征对应的预置规则，对所述待检测原图像进行处理得到检测专用图像；
    所述终端确定常规图像；
    所述终端对所述检测专用图像进行检测，确定检测结果图像；
    所述终端根据所述检测结果图像和所述常规图像，确定待显示图像。
  2. 根据权利要求1所述的方法，其特征在于，所述待检测特征包括：
    肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种，其中，所述眼部特征包括：眼袋、黑眼圈和眼部细纹中的至少一个。
  3. 根据权利要求1或2所述的方法，其特征在于，所述待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
  4. 根据权利要求3所述的方法，其特征在于，所述终端在皮肤检测模式下根据所述待检测特征确定所述待检测原图像，包括：
    当所述待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，所述待检测原图像为所述彩色待检测原图像；
    当所述待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种时，所述待检测原图像为所述黑白待检测原图像。
  5. 根据权利要求3所述的方法，其特征在于，所述终端确定常规图像，包括：
    所述终端融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。
  6. 根据权利要求1所述的方法，其特征在于，所述终端根据所述待检测特征对应的预置规则，对所述待检测原图像进行处理得到检测专用图像，包括：
    当所述待检测特征为肤色时，所述肤色对应的预置规则包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则；
    当所述待检测特征为色斑和红区中的至少一种时，所述色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则；
    当所述待检测特征为痤疮时，所述痤疮对应的预置规则包括：对比度变换、图像锐化和HSV空间图像中的至少一个子规则；
    当所述待检测特征为油分时，所述油分对应的预置规则包括：对比度变换；
    当所述待检测特征为眼部特征和毛孔中的至少一种时，所述眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
  7. 根据权利要求6所述的方法，其特征在于，所述白平衡调整的调整幅度为预设值，所述RBX颜色空间变换和所述HSV空间图像的变换空间为预设值。
  8. 根据权利要求1所述的方法，其特征在于，所述终端根据所述检测结果图像和所述常规图像，确定待显示图像，包括：
    所述终端利用图像匹配和图像融合的方法对所述检测结果图像和所述常规图像进行处理，确定待显示图像。
  9. 根据权利要求1所述的方法，其特征在于，包括：
    所述终端在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。
  10. 根据权利要求1所述的方法，其特征在于，还包括：
    所述终端显示所述待显示图像。
  11. 一种图像处理装置，其特征在于，包括：
    第一确定单元，用于在皮肤检测模式下根据待检测特征确定待检测原图像；
    第二确定单元，用于根据所述待检测特征对应的预置规则，对所述待检测原图像进行处理得到检测专用图像；
    所述第二确定单元还用于，确定常规图像；
    检测单元，用于对检测专用图像进行检测，确定检测结果图像；
    匹配单元，用于根据所述检测结果图像和常规图像，确定待显示图像。
  12. 根据权利要求11所述的装置，其特征在于，所述待检测特征包括：
    肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种，其中，所述眼部特征包括：眼袋、黑眼圈和眼部细纹中的至少一个。
  13. 根据权利要求11或12所述的装置，其特征在于，所述待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
  14. 根据权利要求13所述的装置，其特征在于，包括：
    当所述待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，所述待检测原图像为所述彩色待检测原图像；
    当所述待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种时，所述待检测原图像为所述黑白待检测原图像。
  15. 根据权利要求13所述的装置，其特征在于，所述第二确定单元还用于，融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。
  16. 根据权利要求15所述的装置，其特征在于，包括：
    当所述待检测特征为肤色时，所述肤色对应的预置规则包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则；
    当所述待检测特征为色斑和红区中的至少一种时，所述色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则；
    当所述待检测特征为痤疮时，所述痤疮对应的预置规则包括：对比度变换、图像锐化和HSV空间图像中的至少一个子规则；
    当所述待检测特征为油分时，所述油分对应的预置规则包括：对比度变换；
    当所述待检测特征为眼部特征和毛孔中的至少一种时，所述眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
  17. 根据权利要求16所述的装置，其特征在于，所述白平衡调整的调整幅度为预设值，所述RBX颜色空间变换和所述HSV空间图像的变换空间为预设值。
  18. 根据权利要求11所述的装置，其特征在于，所述匹配单元还用于，利用图像匹配和图像融合的方法对所述检测结果图像和所述常规图像进行处理，确定待显示图像。
  19. 根据权利要求11所述的装置，其特征在于，包括：
    所述第一确定模块还用于，在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。
  20. 根据权利要求11所述的装置，其特征在于，所述装置还包括：
    显示模块，用于显示待显示图像。
  21. 一种终端，其特征在于，包括存储器和处理器：
    所述处理器，用于在皮肤检测模式下根据待检测特征确定待检测原图像；
    所述处理器还用于，根据所述待检测特征对应的预置规则，对所述待检测原图像进行处理得到检测专用图像；
    所述处理器还用于，确定常规图像；
    所述处理器还用于，对所述检测专用图像进行检测，确定检测结果图像；
    所述处理器还用于，根据所述检测结果图像和所述常规图像，确定待显示图像。
  22. 根据权利要求21所述的终端，其特征在于，所述待检测特征包括：
    肤色、色斑、红区、痤疮、油分、皱纹、眼部特征和毛孔中的至少一种。
  23. 根据权利要求21或22所述的终端，其特征在于，所述待检测原图像包括：彩色待检测原图像和黑白待检测原图像中的至少一个。
  24. 根据权利要求23所述的终端，其特征在于，包括：
    当所述待检测特征包括：肤色、色斑、红区和痤疮中的至少一种时，所述待检测原图像为所述彩色待检测原图像；
    当所述待检测特征包括：油分、皱纹、眼部特征和毛孔中的至少一种时，所述待检测原图像为所述黑白待检测原图像。
  25. 根据权利要求23所述的终端，其特征在于，所述处理器还用于，融合常规图像的至少两个子图像，其中，至少两个子图像包括：彩色常规子图像和黑白常规子图像中的至少一种。
  26. 根据权利要求21所述的终端，其特征在于，包括：
    当所述待检测特征为肤色时，所述肤色对应的预置规则包括：光照均衡、白平衡调整和图像平滑中的至少一个子规则；
    当所述待检测特征为色斑和红区中的至少一种时，所述色斑和红区中的至少一种对应的预置规则包括：对比度变换、图像锐化和RBX颜色空间变换中的至少一个子规则；
    当所述待检测特征为痤疮时，所述痤疮对应的预置规则包括：对比度变换、图像锐化和HSV空间图像中的至少一个子规则；
    当所述待检测特征为油分时，所述油分对应的预置规则包括：对比度变换；
    当所述待检测特征为眼部特征和毛孔中的至少一种时，所述眼部特征和毛孔中的至少一种对应的预置规则包括：对比度变换、光照均衡和图像锐化中的至少一个子规则。
  27. 根据权利要求26所述的终端，其特征在于，所述白平衡调整的调整幅度为预设值，所述RBX颜色空间变换和所述HSV空间图像的变换空间为预设值。
  28. 根据权利要求21所述的装置，其特征在于，所述处理器还用于，利用图像匹配和图像融合的方法对所述检测结果图像和所述常规图像进行处理，确定待显示图像。
  29. 根据权利要求21所述的终端，其特征在于，所述处理器还用于，在皮肤检测模式下显示检测特征选项，根据用户选择确定待检测特征。
  30. 根据权利要求21所述的终端，其特征在于，还包括：
    显示屏，用于显示所述待显示图像。
  31. 根据权利要求21所述的终端，其特征在于，所述处理器包括：图像信号处理器。
  32. 根据权利要求21所述的终端，其特征在于，所述处理器包括：图像信号处理器和中央处理器；
    所述图像信号处理器，用于在皮肤检测模式下根据待检测特征确定待检测原图像；
    所述图像信号处理器还用于，根据所述待检测特征对应的预置规则，对所述待检测原图像进行处理得到检测专用图像；
    所述图像信号处理器还用于，确定常规图像；
    所述中央处理器，用于对所述检测专用图像进行检测，确定检测结果图像；
    所述中央处理器还用于，根据所述检测结果图像和所述常规图像，确定待显示图像。
  33. 一种计算机可读存储介质，包括指令，当其在计算机上运行时，使得计算机执行如权利要求1-10任意一项所述的方法。
  34. 一种包含指令的计算机程序产品，当其在计算机上运行时，使得计算机执行如权利要求1-10任意一项所述的方法。
PCT/CN2017/098787 2017-08-24 2017-08-24 一种图像检测的方法、装置及终端 Ceased WO2019037014A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
EP17922433.2A EP3664016B1 (en) 2017-08-24 2017-08-24 Image detection method and apparatus, and terminal
PCT/CN2017/098787 WO2019037014A1 (zh) 2017-08-24 2017-08-24 一种图像检测的方法、装置及终端
CN201780064306.5A CN109844804B (zh) 2017-08-24 2017-08-24 一种图像检测的方法、装置及终端
US16/640,945 US11321830B2 (en) 2017-08-24 2017-08-24 Image detection method and apparatus and terminal

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2017/098787 WO2019037014A1 (zh) 2017-08-24 2017-08-24 一种图像检测的方法、装置及终端

Publications (1)

Publication Number Publication Date
WO2019037014A1 true WO2019037014A1 (zh) 2019-02-28

Family

ID=65438278

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/098787 Ceased WO2019037014A1 (zh) 2017-08-24 2017-08-24 一种图像检测的方法、装置及终端

Country Status (4)

Country Link
US (1) US11321830B2 (zh)
EP (1) EP3664016B1 (zh)
CN (1) CN109844804B (zh)
WO (1) WO2019037014A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110148125A (zh) * 2019-05-21 2019-08-20 苏州大学 基于颜色检测的自适应皮肤油脂检测方法
US20220108445A1 (en) * 2020-10-02 2022-04-07 L'oreal Systems and methods for acne counting, localization and visualization
CN115761473A (zh) * 2021-09-03 2023-03-07 杭州九阳小家电有限公司 烹饪设备的信息提示方法及烹饪设备

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111199171B (zh) * 2018-11-19 2022-09-23 荣耀终端有限公司 一种皱纹检测方法和终端设备
US11100639B2 (en) * 2018-12-11 2021-08-24 H-Skin Aesthetics Co., Ltd. Method for skin examination based on RBX color-space transformation
CN110378304B (zh) * 2019-07-24 2021-11-02 厦门美图之家科技有限公司 皮肤状态检测方法、装置、设备及存储介质
KR20210128838A (ko) * 2020-04-17 2021-10-27 엘지이노텍 주식회사 이미지 처리 장치 및 이미지 처리 방법
CA3191566A1 (en) * 2020-08-11 2022-02-17 Cortina Health, Inc. Systems and methods for using artificial intelligence for skin condition diagnosis and treatment options
CN116964642A (zh) * 2020-10-02 2023-10-27 巴黎欧莱雅 用于痤疮计数、定位和可视化的系统和方法
CN113643227B (zh) * 2020-11-30 2025-02-25 马学召 一种基于深度学习技术的护肤效果检测方法及检测装置
US12028601B2 (en) * 2021-03-30 2024-07-02 Snap Inc. Inclusive camera
CN113128373B (zh) * 2021-04-02 2024-04-09 西安融智芙科技有限责任公司 基于图像处理的色斑评分方法、色斑评分装置及终端设备
CN113592851B (zh) * 2021-08-12 2023-06-20 北京滴普科技有限公司 一种基于全脸图像的毛孔检测方法
CN115379208B (zh) * 2022-10-19 2023-03-31 荣耀终端有限公司 一种摄像头的测评方法及设备
CN119671961B (zh) * 2024-11-28 2025-12-16 广州稀咖科技有限公司 一种基于测油面纸的出油等级分析方法及装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1158779A2 (en) * 2000-05-25 2001-11-28 Eastman Kodak Company Color image reproduction of scenes with preferential color mapping and scene-dependent tone scaling
US20100214421A1 (en) * 2009-02-26 2010-08-26 Di Qu Skin Color Measurement
CN103927718A (zh) * 2014-04-04 2014-07-16 北京金山网络科技有限公司 一种图片处理方法及装置
CN106327537A (zh) * 2015-07-02 2017-01-11 阿里巴巴集团控股有限公司 一种图像预处理方法及装置

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070133017A1 (en) * 2004-02-12 2007-06-14 Hideyuki Kobayashi Image processing apparatus, photographing apparatus, image processing system, image processing method and program
JP5290585B2 (ja) * 2008-01-17 2013-09-18 株式会社 資生堂 肌色評価方法、肌色評価装置、肌色評価プログラム、及び該プログラムが記録された記録媒体
WO2011103576A1 (en) * 2010-02-22 2011-08-25 Canfield Scientific, Incorporated Reflectance imaging and analysis for evaluating tissue pigmentation
US9189679B2 (en) * 2010-06-21 2015-11-17 Pola Chemical Industries, Inc. Age estimation method and sex determination method
WO2012011905A1 (en) 2010-07-22 2012-01-26 The Procter & Gamble Company Methods for improving the appearance of hyperpigmented spot(s) with multiple actives
US20130058543A1 (en) * 2011-09-06 2013-03-07 The Proctor & Gamble Company Systems, devices, and methods for image analysis
US20140378810A1 (en) * 2013-04-18 2014-12-25 Digimarc Corporation Physiologic data acquisition and analysis
TW201540264A (zh) 2014-04-18 2015-11-01 Sony Corp 資訊處理裝置、資訊處理方法、及程式
JP2015232746A (ja) * 2014-06-09 2015-12-24 パナソニックＩｐマネジメント株式会社 皺検出装置および皺検出方法
US10368795B2 (en) * 2014-06-30 2019-08-06 Canfield Scientific, Incorporated Acne imaging methods and apparatus
WO2017094188A1 (ja) * 2015-12-04 2017-06-08 株式会社日立製作所 皮膚糖化検査装置、皮膚糖化検査装置システム及び皮膚糖化検査方法
CN105787929B (zh) * 2016-02-15 2018-11-27 天津大学 基于斑点检测的皮肤疹点提取方法
CN106650215A (zh) * 2016-10-11 2017-05-10 武汉嫦娥医学抗衰机器人股份有限公司 一种基于云平台的肤质检测及个性化评价系统及方法
CN116269262A (zh) * 2016-12-01 2023-06-23 松下知识产权经营株式会社 生物体信息检测装置、生物体信息检测方法及存储介质
CN112533534B (zh) * 2018-08-21 2024-07-16 宝洁公司 用于识别毛孔颜色的方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1158779A2 (en) * 2000-05-25 2001-11-28 Eastman Kodak Company Color image reproduction of scenes with preferential color mapping and scene-dependent tone scaling
US20100214421A1 (en) * 2009-02-26 2010-08-26 Di Qu Skin Color Measurement
CN103927718A (zh) * 2014-04-04 2014-07-16 北京金山网络科技有限公司 一种图片处理方法及装置
CN106327537A (zh) * 2015-07-02 2017-01-11 阿里巴巴集团控股有限公司 一种图像预处理方法及装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3664016A4 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110148125A (zh) * 2019-05-21 2019-08-20 苏州大学 基于颜色检测的自适应皮肤油脂检测方法
US20220108445A1 (en) * 2020-10-02 2022-04-07 L'oreal Systems and methods for acne counting, localization and visualization
US12387319B2 (en) * 2020-10-02 2025-08-12 L'oreal Systems and methods for acne counting, localization and visualization
CN115761473A (zh) * 2021-09-03 2023-03-07 杭州九阳小家电有限公司 烹饪设备的信息提示方法及烹饪设备

Also Published As

Publication number Publication date
EP3664016A4 (en) 2020-07-29
CN109844804A (zh) 2019-06-04
EP3664016B1 (en) 2022-06-22
US11321830B2 (en) 2022-05-03
EP3664016A1 (en) 2020-06-10
US20200380674A1 (en) 2020-12-03
CN109844804B (zh) 2023-06-06

Similar Documents

Publication Publication Date Title
CN109844804B (zh) 一种图像检测的方法、装置及终端
US8320641B2 (en) Method and apparatus for red-eye detection using preview or other reference images
US9691136B2 (en) Eye beautification under inaccurate localization
EP4131065B1 (en) SKIN COLOUR DETECTION METHOD AND APPARATUS, TERMINAL AND STORAGE MECHANISM
CN107993209B (zh) 图像处理方法、装置、计算机可读存储介质和电子设备
CN107798652A (zh) 图像处理方法、装置、可读存储介质和电子设备
CN107742274A (zh) 图像处理方法、装置、计算机可读存储介质和电子设备
CN107911625A (zh) 测光方法、装置、可读存储介质和计算机设备
AU2015201759A1 (en) Electronic apparatus for providing health status information, method of controlling the same, and computer readable storage medium
JP2005086516A (ja) 撮像装置、印刷装置、画像処理装置およびプログラム
CN107862658B (zh) 图像处理方法、装置、计算机可读存储介质和电子设备
CN107862653A (zh) 图像显示方法、装置、存储介质和电子设备
CN109618098A (zh) 一种人像面部调整方法、装置、存储介质及终端
WO2023130922A1 (zh) 图像处理方法与电子设备
JP2019533269A (ja) 人間の強膜および瞳孔に基づいてデジタル画像の色を補正するシステムおよび方法
CN107578372B (zh) 图像处理方法、装置、计算机可读存储介质和电子设备
WO2025189937A9 (zh) 图像处理方法及电子设备
WO2024082976A1 (zh) 文本图像的ocr识别方法、电子设备及介质
JP2018538602A (ja) 毛髪の直径測定
JP2018532205A (ja) 毛髪の縮れ測定
WO2021128593A1 (zh) 人脸图像处理的方法、装置及系统
CN108093170A (zh) 用户拍照方法、装置及设备
CN107845076A (zh) 图像处理方法、装置、计算机可读存储介质和计算机设备
CN117770774A (zh) 一种甲襞微循环图像处理系统、方法及电子设备
CN115937919A (zh) 一种妆容颜色识别方法、装置、设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17922433

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2017922433

Country of ref document: EP

Effective date: 20200305