WO2017199758A1 - 画像処理装置、画像処理方法、撮像装置、プログラム - Google Patents

画像処理装置、画像処理方法、撮像装置、プログラム Download PDF

Info

Publication number
WO2017199758A1
WO2017199758A1 PCT/JP2017/017218 JP2017017218W WO2017199758A1 WO 2017199758 A1 WO2017199758 A1 WO 2017199758A1 JP 2017017218 W JP2017017218 W JP 2017017218W WO 2017199758 A1 WO2017199758 A1 WO 2017199758A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
motion
images
correction
time
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/JP2017/017218
Other languages
English (en)
French (fr)
Inventor
幸直 見城
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.)
Sony Corp
Original Assignee
Sony Corp
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 Sony Corp filed Critical Sony Corp
Priority to US16/087,791 priority Critical patent/US11064129B2/en
Priority to CN201780025394.8A priority patent/CN109076170B/zh
Publication of WO2017199758A1 publication Critical patent/WO2017199758A1/ja
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/70Circuitry for compensating brightness variation in the scene
    • H04N23/73Circuitry for compensating brightness variation in the scene by influencing the exposure time
    • GPHYSICS
    • G03PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
    • G03BAPPARATUS OR ARRANGEMENTS FOR TAKING PHOTOGRAPHS OR FOR PROJECTING OR VIEWING THEM; APPARATUS OR ARRANGEMENTS EMPLOYING ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ACCESSORIES THEREFOR
    • G03B5/00Adjustment of optical system relative to image or object surface other than for focusing
    • GPHYSICS
    • G03PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
    • G03BAPPARATUS OR ARRANGEMENTS FOR TAKING PHOTOGRAPHS OR FOR PROJECTING OR VIEWING THEM; APPARATUS OR ARRANGEMENTS EMPLOYING ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ACCESSORIES THEREFOR
    • G03B15/00Special procedures for taking photographs; Apparatus therefor
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60Control of cameras or camera modules
    • H04N23/68Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
    • H04N23/681Motion detection
    • H04N23/6811Motion detection based on the image signal
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60Control of cameras or camera modules
    • H04N23/68Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
    • H04N23/682Vibration or motion blur correction
    • H04N23/683Vibration or motion blur correction performed by a processor, e.g. controlling the readout of an image memory
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60Control of cameras or camera modules
    • H04N23/68Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
    • H04N23/682Vibration or motion blur correction
    • H04N23/684Vibration or motion blur correction performed by controlling the image sensor readout, e.g. by controlling the integration time
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/70Circuitry for compensating brightness variation in the scene
    • H04N23/741Circuitry for compensating brightness variation in the scene by increasing the dynamic range of the image compared to the dynamic range of the electronic image sensors

Definitions

  • the present technology relates to an image processing device, an image processing method, an imaging device, and a program.
  • the present technology relates to an image processing device, an image processing method, an imaging device, and a program that can more appropriately reduce the influence of motion on image quality.
  • motion correction when correction (hereinafter referred to as motion correction) for reducing the influence of a shift due to motion such as camera shake is performed, it is performed on a recording frame basis.
  • motion correction when synthesizing images captured in time division with different sensitivities, motion correction is performed with reference to the synthesized image.
  • the correction accuracy may be lowered by performing the motion correction without considering the ratio of the images captured with different sensitivities.
  • This technology has been made in view of such a situation, and makes it possible to more appropriately perform correction for reducing the influence of movement.
  • An image processing apparatus includes a combining unit that combines a plurality of images captured at different timings, and a motion correction unit that performs correction to reduce the influence of the motion on the image.
  • the unit sets a standard of a section that refers to motion amount information related to a motion amount supplied at a predetermined interval based on a composition ratio when the plurality of images are combined, and based on the set criterion, The correction is performed with reference to the motion amount information.
  • An image processing method includes a step of combining a plurality of images captured at different timings and performing motion correction to reduce the influence of motion on the image, and the motion correction is performed at a predetermined interval.
  • the reference of the section that refers to the motion amount information related to the amount of motion supplied in step S is set based on the combination ratio when combining the plurality of images, and the motion amount information is referred to based on the set reference. Processing to be performed.
  • An imaging apparatus includes a synthesis unit that synthesizes a plurality of images captured at different timings, and a motion correction unit that performs correction to reduce the influence of motion on the image.
  • a synthesis unit that synthesizes a plurality of images captured at different timings
  • a motion correction unit that performs correction to reduce the influence of motion on the image.
  • a program includes a step of combining a plurality of images captured at different timings with a computer to perform motion correction that reduces an influence on motion with respect to the image.
  • a standard for a section that refers to motion amount information related to the amount of motion supplied at intervals is set based on a combination ratio when the plurality of images are combined, and the motion amount information is referenced based on the set reference. The process to be performed is executed.
  • a plurality of images captured at different timings are combined, and motion correction is performed to reduce the influence of the motion on the image.
  • a section reference that refers to motion amount information related to a motion amount supplied at a predetermined interval is set based on a combination ratio when a plurality of images are combined. This is done with reference to the quantity information.
  • the imaging apparatus includes the image processing apparatus.
  • image processing device and the imaging processing device may be independent devices or may be internal blocks constituting one device.
  • the program can be provided by being transmitted through a transmission medium or by being recorded on a recording medium.
  • correction for reducing the influence of movement can be performed more appropriately.
  • the present technology described below can be applied to an apparatus that performs correction (hereinafter, referred to as motion correction) for reducing the influence of shift (blur) due to motion. Further, the present invention can be applied to an apparatus that captures an image with a wide dynamic range, performs motion correction on the image with such a wide dynamic range, and improves image quality.
  • motion correction for reducing the influence of shift (blur) due to motion.
  • the present invention can be applied to an apparatus that captures an image with a wide dynamic range, performs motion correction on the image with such a wide dynamic range, and improves image quality.
  • a wide dynamic range image can be generated by synthesizing the long exposure image and the short exposure image.
  • the wide dynamic range image generation process will be described with reference to FIG.
  • FIG. 1 shows an example of processing for generating a wide dynamic range (HDR) image 55 by combining (blending) the short-time exposure image 51 and the long-time exposure image 52.
  • the synthesis (blending) process in the process of the present disclosure is a blending process of corresponding pixel values of the short-time exposure image 51 and the long-time exposure image 52 according to the blend coefficient ⁇ . Each pixel value of (image) is set.
  • exposure correction is performed on the short-time exposure image 51 and the long-time exposure image 52 by multiplying by a constant corresponding to the exposure ratio in Step S11 and Step S12.
  • a constant corresponding to the exposure ratio For example, in the case of a 10-bit sensor with an exposure ratio of 16, the short-time exposure image is multiplied by 16 at the exposure correction stage, the long-time exposure image is multiplied by 1, and then the pixel values of the corresponding pixels are synthesized (blended) in step S13.
  • the pixel value of the HDR image 55 as the output image is determined.
  • the short exposure image after exposure correction is referred to as an exposure correction short exposure image
  • the long exposure image after exposure correction is referred to as an exposure correction long exposure image
  • DS Pixel value of exposure-corrected short-exposure image
  • DL Pixel value of exposure-corrected long-exposure image
  • DH Pixel value of HDR image 55 when the pixel value of the wide dynamic range image to be output is blended in step S13 Is calculated according to the following formula.
  • pixel value blending according to the above equation is performed for each corresponding pixel position in the exposure-corrected short-time exposure image and the exposure-corrected long-time exposure image. That is, blend processing is performed for each imaging pixel position of the same subject, and each pixel value of the output image (HDR image) is determined.
  • the exposure-corrected short-time exposure image 53 and the exposure-corrected long-time exposure image 54 obtained by performing exposure correction by multiplying by a constant corresponding to the exposure ratio in steps S11 and S12 hold accurate values as pixel values. It may not be. That is, there is a possibility that the dark portion of the short-time exposure image 53 is crushed by noise, and the bright portion of the long-time exposure image 54 is whiteout. Therefore, in the blend process in step S13, a blend process that does not use the pixel values of these error pixels is performed. By such processing, an HDR image 55 having a good SN ratio (SNR) can be obtained from the bright part to the dark part.
  • SNR good SN ratio
  • FIG. 2 is a diagram illustrating a configuration of an embodiment of an imaging apparatus to which the present technology is applied.
  • the imaging apparatus 100 illustrated in FIG. 2 is an apparatus that generates the HDR image 55 by performing the processing described with reference to FIG.
  • image pickup apparatus 100 In the image pickup apparatus 100 shown in FIG. 2, light incident through the optical lens 101 enters an image pickup device 102 constituted by an image pickup unit, for example, a CMOS image sensor, and outputs image data by photoelectric conversion.
  • image data is input to the image processing unit 103.
  • the output image of the image sensor 102 is a so-called mosaic image in which any pixel value of RGB is set for each pixel.
  • the image processing unit 103 generates a high dynamic range (HDR) image based on the demosaic process that sets all RGB pixel values for each pixel and the above-described synthesis process of the long exposure image and the short exposure image. Processing, motion correction processing described later, and the like are performed.
  • HDR high dynamic range
  • a sensor such as an acceleration sensor or a gyro sensor
  • a sensor can be configured to be provided in the image processing unit 103 or can be configured to be provided outside the image processing unit 103.
  • the output from the image processing unit 103 is input to the signal processing unit 104.
  • the signal processing unit 104 performs signal processing in a general camera, such as white balance (WB) adjustment and gamma correction, and generates an output image 120.
  • the output image 120 is stored in a storage unit (not shown). Or it outputs to a display part (not shown).
  • the control unit 105 outputs a control signal to each unit according to a program stored in a memory (not shown), for example, and controls various processes.
  • FIG. 3 shows the configuration of the image processing unit 103.
  • the image processing unit 103 performs a wide dynamic range image generation process, a motion correction process, and the like.
  • the pre-correction image 151 from the image sensor 102 is input to the image processing unit 103.
  • the pre-correction image 151 is supplied to the HDR processing unit 131.
  • the HDR processing unit 131 generates an HDR image 55 from the supplied pre-correction image 151 and supplies it to the motion correction unit 132.
  • the pre-correction image 151 is a long-time exposure image and a short-time exposure image
  • the HDR processing unit 131 generates an HDR image 55 from the supplied long-time exposure image and short-time exposure image and sends it to the motion correction unit 132. Supply.
  • the HDR processing unit 131 When the HDR processing unit 131 generates the HDR image 55 (generates a composite image), the HDR processing unit 131 also generates a composite statistic 152 and supplies it to the motion correction unit 132. As will be described later with reference to FIG. 4, the total statistic 152 is a statistic related to the time referring to the long exposure image and the short exposure image when the HDR image 55 is generated.
  • the total statistic 152 can be a histogram result or a reference time centroid obtained from the histogram.
  • the histogram result may be supplied to the motion correction unit 132 as the combined statistic 152, and the reference time centroid may be obtained from the histogram by the motion correction unit 132. It may be supplied to the motion correction unit 132.
  • the motion correction unit 132 sets motion amount information to be referred to based on the combined statistics 152 for the supplied HDR image 55, and based on the referenced motion amount information, a correction amount for suppressing the influence of motion Is calculated and corrected.
  • the motion correction unit 132 outputs the corrected image 153 subjected to the motion correction to the signal processing unit 104 (FIG. 2).
  • FIG. 4 is a diagram illustrating a relationship among an imaging frame, a recording frame, a composite image reference time histogram, motion amount information, and a reference range.
  • the imaging frame includes a long exposure image and a short exposure image, and a single recording frame is generated by combining the long exposure image and the short exposure image.
  • the time t1 to the time t3 are a long exposure period, and an image captured by the image sensor 102 within the long exposure period is set as the long exposure image 1L in the image processing unit 103 (the HDR processing unit 131 therein). ).
  • time t2 to the time t4 are set as a short exposure period, and an image captured by the image sensor 102 within the short exposure period is set as the short exposure image 1S in the image processing unit 103 (the HDR processing unit 131 therein). ).
  • a recording frame 1F (HDR image 55) is generated from the long exposure image 1L and the short exposure image 1S obtained from time t1 to time t4.
  • a recording frame 2F (HDR image 55) is generated from the long exposure image 2L and the short exposure image 2S obtained from time t4 to time t7, and the long time obtained from time t7 to time t10.
  • a recording frame 3F (HDR image 55) is generated from the exposure image 3L and the short exposure image 3S. From the long exposure image 4L and the short exposure image 4S obtained from time t10 to time t13, the recording frame 4F ( An HDR image 55) is generated.
  • the recording frame 1F (HDR image 55) is generated from the long exposure image 1L and the short exposure image 1S obtained from time t1 to time t4, as described with reference to FIG.
  • the long exposure image 1L and the short exposure image 1S are synthesized at a predetermined ratio by the blend coefficient ⁇ .
  • the blend coefficient ⁇ is calculated by the following equation.
  • (VS) / (VS + VL)
  • VL Noise variance value of the pixel of interest in the long exposure image after exposure correction
  • VS Noise variance value of the pixel of interest in the short exposure image after exposure correction
  • Correction is performed so as to be zero.
  • the blend coefficient ⁇ may be different for each pixel (depending on the setting such as for each frame or for each line in the frame).
  • the composition ratio of the long exposure image and the short exposure image may be different for each pixel and is not constant.
  • FIG. 5 shows that the data acquired during the long exposure period and the data acquired during the short exposure time are respectively developed in the frame memory, and the developed long exposure image 1L and short exposure image 1S are developed.
  • FIG. 6 is a diagram showing a part (pixel arrangement) of a long exposure image 1L, a short exposure image 1S, and a recording frame 1F when a recording frame 1F is generated.
  • the pixel PF1 of the recording frame 1F is generated from the pixel PL1 of the long exposure image 1L and the pixel PS1 of the short exposure image 1S.
  • the pixel PL1 of the long exposure image 1L, the pixel PS1 of the short exposure image 1S, and the pixel PF1 of the recording frame 1F are pixels at the same position in each image.
  • a pixel PF2 of the recording frame 2F is generated from the pixel PL2 of the long exposure image 2L and the pixel PS2 of the short exposure image 2S.
  • the blend ratio of the pixel PL1 and the pixel PS1 is set by the blend coefficient ⁇ described above. Is done. It is assumed that the blend ratio of the pixel PL1 and the pixel PS1 is, for example, 2: 1.
  • the blend ratio of the pixel PL2 and the pixel PS2 is, for example, 3: 1.
  • the HDR processing unit 131 sets a blend coefficient ⁇ , and combines the pixel PL of the long exposure image L and the pixel PS of the short exposure image S based on the blend coefficient ⁇ , so that the HDR image 55 (record frame F ) Pixel PF. Further, the HDR processing unit 131 generates the HDR image 55 in this way, and also generates a composite image reference time histogram (FIG. 4).
  • the horizontal axis represents time
  • the vertical axis represents frequency.
  • the frequency can be, for example, a blend ratio (blend coefficient ⁇ ).
  • the pixel values PL1 and PL2 of the long-exposure image 1L are obtained at the same time (time tL), and the pixels PS1 and PS2 of the short-time exposure image 1S are the same time (time tS). ) Is the acquired pixel value.
  • the blend ratio of the pixel PL1 and the pixel PS1 is 2: 1
  • the blend ratio of the pixel PL2 and the pixel PS2 is 3: 1.
  • a value (blend ratio) for each time is calculated, and a composite image reference time histogram is generated.
  • the blend ratio value obtained from the blend coefficient ⁇
  • a histogram can be created using other values.
  • the motion correction amount is calculated using the motion amount information acquired in the time (section) from this reference time centroid to the next reference time centroid. For example, the time T1 from the composite image reference time histogram when the recording frame 1F is processed is set as the reference time centroid, and the time T2 from the composite image reference time histogram when the recording frame 2F is processed is the reference time centroid. Is set to motion amount information obtained in a section from time T1 to time T2 as motion amount information referred to in the recording frame 2F, and a motion correction amount is calculated.
  • the movement amount information is output at predetermined intervals (cycles) from sensors such as an acceleration sensor and a gyro sensor, for example. Therefore, in this case, if the section becomes longer, the amount of motion information referred to increases, and if the section becomes shorter, the amount of motion information referred to decreases. In addition, when the reference time center of gravity is closer to the long exposure period, the motion amount information obtained within the long exposure period is dominant, and when it is closer to the short exposure period, the movement amount information is shorter. The amount of motion information obtained during the exposure period becomes dominant.
  • the composite image reference time histogram obtained from the recording frame 4F from the composite image reference time histogram obtained from the recording frame 1F is gradually determined from the long-time exposure image L. And the histogram obtained from the short-time exposure image S is large. Along with such a change, the reference time center of gravity has also moved to the histogram side obtained from the short-time exposure image S.
  • a section of information to be referred to for correcting a blur amount due to a motion is set based on the combination ratio. . Therefore, more appropriate correction can be performed.
  • FIG. 6 shows an imaging frame and a recording frame as in FIG. 4, and is a diagram for comparing the case where the reference time centroid is calculated as described above and the case where the reference time centroid is calculated by a conventional method.
  • C1 represents a long exposure period
  • C2 represents a short exposure period
  • B2 represents the exposure interval between the long exposure period and the short exposure period.
  • T1 ′ represents the reference time centroid calculated by the conventional method
  • the time without the dash is a reference time calculated by applying the present technology. Represents the center of gravity.
  • the recording frame rate is 30 (ftp)
  • the long exposure period C1 is 31 (msec)
  • the short exposure period C2 is 0.1 (msec)
  • the exposure interval B2 is 1 ( msec).
  • the blend ratio is set to use all the short-time exposure images and not to blend the long-time exposure images.
  • the reference time center of gravity is set on the short-exposure image side.
  • time T1, time T2, time T3, and time T4 in FIG. 6 the reference time centroid is set on the short exposure image side, specifically, at the central time of the short exposure period.
  • the reference time center of gravity for example, the time at the center of one recording frame, in other words, the long exposure period C1, the short exposure period C2, and the exposure interval.
  • the central time obtained by adding B2 is set as the reference time centroid.
  • the reference time centroid is set at a substantially central time of the long exposure period.
  • the reference time centroid obtained by the present embodiment and the reference time centroid obtained by the conventional method may differ greatly.
  • the motion amount information obtained when the short-exposure image is captured is dominant and the motion correction is performed.
  • the motion amount information becomes dominant when a short-time exposure image is captured, so that appropriate motion correction can be performed.
  • FIG. 7 shows the result of simulating the deviation of the reference time center of gravity and the effect of motion correction.
  • the horizontal axis represents the motion frequency (Hz), and the vertical axis represents the remaining correction amount.
  • FIG. 7 shows graphs every 2, 4, 8, 12, 16, 20, 24, 28, and 32 (msec) as the deviation of the reference time center of gravity.
  • a line with a correction amount of 100% has no correction effect (0%), and if it is 100% or more, it is corrected in the reverse direction (incorrect correction). Show. It can be seen that when the deviation of the reference time center of gravity increases, and when the motion frequency increases, it easily exceeds the line where the correction effect is 0%.
  • an appropriate reference time center of gravity can be set (because the deviation of the reference time center of gravity can be made substantially zero), it is possible to perform appropriate correction. .
  • FIG. 8 is a diagram showing the relationship among the imaging frame, recording frame, composite image reference time histogram, motion amount information, and reference range, as in FIG.
  • a composite image reference time histogram is generated on a line y indicating a line in the horizontal direction (y-axis direction) of the imaging frame (recording frame).
  • a histogram is generated for each line, a reference time center of gravity is obtained for each line based on the histogram, a reference range of motion amount information is set, a motion correction amount is calculated, Motion compensation is performed.
  • line y is a line that was exposed between time t1 ′ and time t2 ′ within the long exposure time, and was exposed between time t3 ′ and time t4 ′ within the short exposure time.
  • the histogram relating to the line y in the long exposure image is created by referring to the blend ratio set between the time t1 'and the time t2'. Further, the histogram relating to the line y in the short-time exposure image is created with reference to the blend ratio set between time t3 'and time t4'.
  • the histogram created in this way is used to set the reference time centroid on line y in the recording frame.
  • processing can be performed in the same manner as in the first embodiment performed for each frame, and thus detailed description thereof is omitted.
  • the reference time center of gravity is obtained in consideration of the blend ratio between the long-time exposure image and the short-time exposure image.
  • how to obtain the reference time centroid will be described.
  • the case where the times t1 to t4 (long exposure time) in FIG. 4 are 16 (msec) will be described as an example.
  • T ⁇ 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16 ⁇ (1)
  • 16 (msec) the case of 16 (msec) is illustrated, but the number of elements can be N.
  • each element indicates an elapsed time from Vsync.
  • the first element corresponds to a section in which the elapsed time from Vsync is 0 to 1 msec.
  • the accuracy of the histogram calculation is 1 msec, but is not limited to this accuracy.
  • a [y] [x] reference time identification matrix of pixel [y] [x] (2)
  • the first element is the reference time identification matrix of the long exposure image at the pixel [y] [x]
  • the second element is the reference time of the short exposure image at the pixel [y] [x]. It is an identification matrix.
  • the numerical value in the reference time identification matrix is identification information that is 1 when referring and 0 when not referring.
  • B [y] [x] ⁇ 0.8, 0.2 ⁇
  • the composite image reference time histogram hist [y] of the same line (y coordinate) is defined as follows.
  • B [y] [x] is ⁇ 0.8, 0.2 ⁇ regardless of x, and exemplifies a case where the same composition rate is obtained in the same row.
  • a histogram is generated for each line.
  • a histogram for each line is generated in Expression (4), and the following Expression (5) is used.
  • the reference time centroid for each line is calculated.
  • the time centroid center (hist [y]) of the composite image reference time histogram, that is, the reference time centroid is obtained by the following equation.
  • the histogram is created by the following equation.
  • hist [y] is a histogram calculated by equation (4).
  • Expression (6) is an expression for generating a histogram of a frame by integrating the histogram (horizontal histogram) for each line shown in Expression (4) in the vertical direction.
  • the time centroid center (hist) of the composite image reference time histogram that is, the reference time centroid is obtained by the following expression.
  • Equation (7) the time centroid of the composite image reference time histogram of the entire frame, that is, the reference time centroid in one frame is obtained.
  • the reference time centroid is obtained after the histogram is created, but the step of creating the histogram is omitted and the reference time centroid is directly obtained. You may be allowed to.
  • equation (4) the step of obtaining the composite image reference time histogram is omitted, and the time centroid of the composite image reference time histogram is expressed by equation (5) (or equation (7)). It may be calculated.
  • FIG. 9 is a diagram illustrating another configuration of the image processing unit 103.
  • the image processing unit 103 illustrated in FIG. 9 performs a wide dynamic range image generation process, a motion correction process, and the like, similar to the image processing unit 103 illustrated in FIG.
  • the pre-correction image 231 from the image sensor 102 is input to the image processing unit 103 illustrated in FIG.
  • the pre-correction image 231 is supplied to the motion correction unit 201.
  • the motion correction unit 201 performs motion correction on the supplied pre-correction image 231.
  • the combined statistic 234 output from the HDR processing unit 202 is supplied to the motion correction unit 201 via the feedback unit 203, and the motion correction unit 201 performs the pre-correction image 231 based on the supplied combined statistic 234. Motion correction is performed.
  • the pre-correction image 231 is a long-time exposure image and a short-time exposure image
  • the motion correction unit 201 performs motion correction on each of the long-time exposure image and the short-time exposure image.
  • the post-correction image 232 that has been subjected to motion correction by the motion correction unit 201 is supplied to the HDR processing unit 202.
  • the HDR processing unit 202 generates an HDR image 233 from the supplied post-correction image 232 (for example, a long-exposure image and a short-exposure image subjected to motion correction), and sends them to the signal processing unit 104 (FIG. 2). Supply.
  • the HDR processing unit 202 When generating the HDR image 233 (generating a composite image), the HDR processing unit 202 also generates a composite statistic 234 and supplies it to the motion correction unit 201 via the feedback unit 203.
  • the total statistic 234 is a statistic (histogram or histogram) related to the time referring to the long exposure image and the short exposure image when the HDR image 233 is generated.
  • the reference time center of gravity obtained from (1).
  • the image processing unit 103 illustrated in FIG. 3 is a processing flow for performing motion correction after generating an HDR image.
  • the image processing unit 103 illustrated in FIG. 9 generates an HDR image after performing motion correction.
  • the processing flow is different from the image processing unit 103 shown in FIG.
  • the image processing unit 103 illustrated in FIG. 9 is a processing flow for generating an HDR image after performing motion correction.
  • this is a processing flow in which the composite statistic 234 is fed back by the feedback unit 203. Therefore, in other words, since the (n + 1) th frame is generated using the composite statistic 234 when the nth frame is converted into the HDR image, the motion correction is performed with the composite statistic 234 delayed by 1V. Will be done.
  • the composite statistic 234 is not information that varies greatly from frame to frame, the accuracy does not drop significantly even if such a delay occurs.
  • the present technology it is possible to perform motion correction based on the time at which the composite image is captured, and it is possible to prevent the correction accuracy of motion correction from being lowered due to the composite result, and the effect of motion correction can be reduced to 100. It can be demonstrated in a state close to%.
  • This technology can be applied when combining multiple images shot at different shooting timings.
  • the plurality of images are two images of a long-exposure image and a short-exposure image.
  • the present technology can be applied even when two or more images are combined. it can.
  • the present technology can also be applied when noise reduction processing is performed by combining a plurality of images shot at different shooting timings. That is, the present technology can be applied to noise reduction processing in addition to motion correction.
  • a pixel that captures a long-exposure image and a pixel that captures a short-exposure image are arranged in a pixel array unit (not shown) and a long-exposure image is captured (during a long-exposure period).
  • the present technology can also be applied when a short-exposure image is taken (there is a short exposure period). Even when long-exposure images and short-exposure images are captured by spatial division, the long-exposure image and the short-exposure image are images that are captured at different timings.
  • the present technology can be applied to the case of synthesis.
  • the reference time centroid is set according to the combination ratio (blend ratio) when images captured at different timings are combined, and the section for referring to the motion amount information is set.
  • the processing using the motion amount information referred to by setting such an appropriate section may be correction (processing) other than the motion correction described above, and performs processing other than motion correction. Even in this case, the accuracy of the processing can be improved.
  • the series of processes described above can be executed by hardware or can be executed by software.
  • a program constituting the software is installed in the computer.
  • the computer includes, for example, a general-purpose personal computer capable of executing various functions by installing various programs by installing a computer incorporated in dedicated hardware.
  • FIG. 10 is a block diagram showing an example of the hardware configuration of a computer that executes the above-described series of processing by a program.
  • a CPU Central Processing Unit
  • ROM Read Only Memory
  • RAM Random Access Memory
  • An input / output interface 505 is further connected to the bus 504.
  • An input unit 506, an output unit 507, a storage unit 508, a communication unit 509, and a drive 510 are connected to the input / output interface 505.
  • the input unit 506 includes a keyboard, a mouse, a microphone, and the like.
  • the output unit 507 includes a display, a speaker, and the like.
  • the storage unit 508 includes a hard disk, a nonvolatile memory, and the like.
  • the communication unit 509 includes a network interface or the like.
  • the drive 510 drives a removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory.
  • the CPU 501 loads the program stored in the storage unit 508 to the RAM 503 via the input / output interface 505 and the bus 504 and executes the program, for example. Is performed.
  • the program executed by the computer (CPU 501) can be provided by being recorded in, for example, a removable medium 511 as a package medium or the like.
  • the program can be provided via a wired or wireless transmission medium such as a local area network, the Internet, or digital satellite broadcasting.
  • the program can be installed in the storage unit 508 via the input / output interface 505 by attaching the removable medium 511 to the drive 510. Further, the program can be received by the communication unit 509 via a wired or wireless transmission medium and installed in the storage unit 508. In addition, the program can be installed in the ROM 502 or the storage unit 508 in advance.
  • the program executed by the computer may be a program that is processed in time series in the order described in this specification, or in parallel or at a necessary timing such as when a call is made. It may be a program for processing.
  • system represents the entire apparatus composed of a plurality of apparatuses.
  • this technique can also take the following structures.
  • a combining unit that combines a plurality of images captured at different timings;
  • a motion correction unit that performs correction to reduce the influence of the motion on the image,
  • the motion correction unit is A standard of a section that refers to motion amount information related to a motion amount supplied at a predetermined interval is set based on a combination ratio when combining the plurality of images,
  • An image processing apparatus that performs the correction by referring to the motion amount information based on the set standard.
  • the plurality of images are a long exposure image captured in a long exposure period and a short exposure image captured in a short exposure period, The image processing apparatus according to (1), wherein the combining unit generates an image with a wide dynamic range by combining the long-time exposure image and the short-time exposure image at the combination ratio.
  • the plurality of images are a plurality of images captured at the same exposure time and captured at different timings, The image processing device according to (1), wherein the combining unit generates an image with reduced noise by combining the plurality of images.
  • the motion correction unit is A combined image generated by combining a plurality of images by the combining unit is a histogram related to a time at which the plurality of images are referred to as a statistic.
  • a standard of a section that refers to motion amount information related to a motion amount supplied at a predetermined interval is set based on a combination ratio when combining the plurality of images,
  • An image processing method including processing performed with reference to the motion amount information based on the set standard.
  • a combining unit that combines a plurality of images captured at different timings;
  • a motion correction unit that performs correction to reduce the influence of the motion on the image,
  • the motion correction unit is A standard of a section that refers to motion amount information related to a motion amount supplied at a predetermined interval is set based on a combination ratio when combining the plurality of images,
  • An imaging apparatus comprising: an image processing apparatus that performs the correction with reference to the motion amount information based on the set reference.
  • a standard of a section that refers to motion amount information related to a motion amount supplied at a predetermined interval is set based on a combination ratio when combining the plurality of images, A program for executing a process performed with reference to the motion amount information based on the set standard.
  • imaging device 101 optical lens, 102 imaging device, 103 image processing unit, 104 signal processing unit, 105 control unit, 131 HDR processing unit, 132 motion correction unit, 201 motion correction unit, 202 HDR processing unit, 203 feedback unit

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Studio Devices (AREA)
  • Adjustment Of Camera Lenses (AREA)
  • Image Processing (AREA)

Abstract

本技術は、動き補正をより適切に行えるようにする画像処理装置、画像処理方法、撮像装置、プログラムに関する。 異なるタイミングで撮像された複数の画像を合成する合成部と、動きに関する画像への影響を低減する補正を行う動き補正部とを備え、動き補正部は、所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、複数の画像を合成するときの合成比率に基づき設定し、設定された基準を元に、動き量情報を参照し、補正を行う。複数の画像は、長時間露光期間で撮像された長時間露光画像と、短時間露光期間で撮像された短時間露光画像であり、合成部は、長時間露光画像と短時間露光画像を合成比率で合成することで、ダイナミックレンジの広い画像を生成する。本技術は、撮像装置に適用できる。

Description

画像処理装置、画像処理方法、撮像装置、プログラム
 本技術は画像処理装置、画像処理方法、撮像装置、プログラムに関し、例えば、動きによる画質への影響をより適切に低減できるようにした画像処理装置、画像処理方法、撮像装置、プログラムに関する。
 従来、撮像装置のダイナミックレンジ拡大技術が存在する。例えば、異なる感度で時分割に撮像し、時分割に撮像した複数の画像を合成する時分割方式が知られている。(例えば、特許文献1参照)
特開2001-346096号公報
 例えば手振れといった動きによるずれの影響を低減するための補正(以下、動き補正と記述する)が行われる場合、記録フレーム基準で行われる。上記したように、異なる感度で時分割に撮像した画像を合成するような場合、合成後の画像を基準として動き補正が行われる。しかしながら、異なる感度で撮像された画像をどのような割合で合成したかを考慮せずに動き補正が行われることで、補正精度が低下してしまう可能性があった。
 本技術は、このような状況に鑑みてなされたものであり、動きによる影響を低減するための補正をより適切にできるようにするものである。
 本技術の一側面の画像処理装置は、異なるタイミングで撮像された複数の画像を合成する合成部と、動きに関する前記画像への影響を低減する補正を行う動き補正部とを備え、前記動き補正部は、所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、前記設定された基準を元に、前記動き量情報を参照し、前記補正を行う。
 本技術の一側面の画像処理方法は、異なるタイミングで撮像された複数の画像を合成し、動きに関する前記画像への影響を低減する動き補正を行うステップを含み、前記動き補正を、所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、前記設定された基準を元に、前記動き量情報を参照して行う処理を含む。
 本技術の一側面の撮像装置は、異なるタイミングで撮像された複数の画像を合成する合成部と、動きに関する前記画像への影響を低減する補正を行う動き補正部とを備え、前記動き補正部は、所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、前記設定された基準を元に、前記動き量情報を参照し、前記補正を行う画像処理装置を備える。
 本技術の一側面のプログラムは、コンピュータに、異なるタイミングで撮像された複数の画像を合成し、動きに関する前記画像への影響を低減する動き補正を行うステップを含み、前記動き補正を、所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、前記設定された基準を元に、前記動き量情報を参照して行う処理を実行させる。
 本技術の一側面の画像処理装置、画像処理方法、並びにプログラムにおいては、異なるタイミングで撮像された複数の画像が合成され、動きに関する画像への影響を低減する動き補正が行われる。その動き補正は、所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準が、複数の画像を合成するときの合成比率に基づき設定され、設定された基準を元に、動き量情報が参照されて行われる。
 本技術の一側面の撮像装置においては、前記画像処理装置が備えられている。
 なお、画像処理装置および撮像処理装置は、独立した装置であっても良いし、1つの装置を構成している内部ブロックであっても良い。
 また、プログラムは、伝送媒体を介して伝送することにより、または、記録媒体に記録して、提供することができる。
 本技術の一側面によれば、動きによる影響を低減するための補正をより適切にできる。
 なお、ここに記載された効果は必ずしも限定されるものではなく、本開示中に記載されたいずれかの効果であってもよい。
広ダイナミックレンジ画像の生成処理について説明するための図である。 本技術を適用した撮像装置の構成について説明するための図である。 画像処理部の構成について説明するための図である。 フレーム毎に参照時刻重心を求める場合について説明するための図である。 合成について説明するための図である。 参照時刻重心のずれについて説明するための図である。 補正の効果について説明するための図である。 ライン毎に参照時刻重心を求める場合について説明するための図である。 画像処理部の他の構成について説明するための図である。 記録媒体について説明するための図である。
 以下に、本技術を実施するための形態(以下、実施の形態という)について説明する。以下に説明する本技術は、動きによるずれ(ブレ)の影響を低減するための補正(以下、動き補正と記述する)を行う装置に適用できる。また、広ダイナミックレンジの画像を撮像し、そのような広ダイナミックレンジの画像に対して動き補正を行い、画質を向上させる装置に対して適用できる。
 また、露光時間の等しい複数回の撮像(露光時間は等しいが異なるタイミングでの撮像)を行い、位置合わせと合成処理を行うことによりノイズを低減させる処理を行い、画質を向上させる装置に対しても適用できる。
 ここでは、広ダイナミックレンジの画像を撮像し、そのような広ダイナミックレンジの画像に対して動き補正を行う場合を例に挙げて説明する。
 <広ダイナミックレンジ画像の生成処理の概要について>
 まず、広ダイナミックレンジ画像の生成処理の概要について説明する。長時間露光画像と、短時間露光画像を合成することにより広ダイナミックレンジ画像を生成することができる。図1を参照して広ダイナミックレンジ画像の生成処理について説明する。
 図1は、短時間露光画像51と、長時間露光画像52の合成(ブレンド)を行って広ダイナミックレンジ(HDR:High Dynamic Range)画像55を生成する処理例を示している。本開示の処理における合成(ブレンド)処理とは、ブレンド係数βに応じた短時間露光画像51と、長時間露光画像52の対応画素値のブレンド処理であり、このブレンド処理によって、出力画像(合成画像)の各画素値を設定する。
 まず、短時間露光画像51と、長時間露光画像52に対して、ステップS11、ステップS12において、露光比に応じた定数を乗算して露出補正が行われる。例えば10bitのセンサ、露光比16の場合、露出補正の段階で短時間露光画像は16倍し、長時間露光画像は1倍して、その後、ステップS13で対応画素の画素値を合成(ブレンド)して出力画像であるHDR画像55の画素値が決定される。
 なお、露出補正後の短時間露光画像を露出補正短時間露光画像と称し、露出補正後の長時間露光画像を露出補正長時間露光画像と称す。
 出力画像であるHDR画像55の画素値決定処理の具体的シーケンスについて説明する。
 例えば、
 DS:露出補正短時間露光画像の画素値
 DL:露出補正長時間露光画像の画素値
 DH:出力する広ダイナミックレンジ画像の画素値
としたとき、HDR画像55の画素値は、ステップS13のブレンド処理において、以下に示す式に従って算出される。
 DH=(1.0-β)×DS+β×DL
 なお、上記式に従った画素値のブレンドは、露出補正短時間露光画像と、露出補正長時間露光画像の対応画素位置毎に行われる。すなわち同一被写体の撮像画素位置毎にブレンド処理が行われ、出力画像(HDR画像)の各画素値が決定される。
 このような処理により、例えば露出補正短時間露光画素の飽和画素値は1023×16=16368(14bit)となるため、出力する画素値も14bitとなりダイナミックレンジの拡張が実現される。
 ただし、ステップS11,S12において露光比に応じた定数を乗算して露出補正を行って得られる露出補正短時間露光画像53と露出補正長時間露光画像54は、画素値として正確な値を保持していない可能性がある。すなわち、短時間露光画像53の暗部はノイズでつぶれ、長時間露光画像54の明部は白飛びしている可能性がある。そこで、ステップS13におけるブレンド処理に際しては、これらのエラー画素の画素値については利用しないブレンド処理が行われる。このような処理によって明部から暗部までSN比(SNR)の良いHDR画像55が得られる。
 <撮像装置の構成>
 図2は、本技術を適用した撮像装置の一実施の形態の構成を示す図である。また、図2に示した撮像装置100は、図1を参照して説明した処理を行うことで、HDR画像55を生成する装置である。
 図2に示した撮像装置100は、光学レンズ101を介して入射される光は撮像部、例えばCMOSイメージセンサなどによって構成される撮像素子102に入射し、光電変換による画像データを出力する。出力画像データは画像処理部103に入力される。
 撮像素子102の出力画像は、各画素にRGBのいずれかの画素値が設定されたいわゆるモザイク画像である。画像処理部103は、各画素にRGBの全画素値を設定するデモザイク処理、上述した長時間露光画像と短時間露光画像との合成処理に基づく広ダイナミックレンジ(HDR:High Dynamic Range)画像の生成処理、後述する動き補正処理などを行う。
 動き補正処理には、加速度センサやジャイロセンサといったセンサから、周期的に出される動き量情報が用いられる。このようなセンサ(不図示)は、画像処理部103内に備えられている構成とすることも可能であるし、画像処理部103外に備えられている構成とすることも可能である。
 画像処理部103からの出力は、信号処理部104に入力される。信号処理部104は、例えばホワイトバランス(WB)調整、ガンマ補正等、一般的なカメラにおける信号処理を実行して出力画像120を生成する。出力画像120は図示しない記憶部に格納される。あるいは表示部(不図示)に出力される。
 制御部105は、例えば図示しないメモリに格納されたプログラムに従って各部に制御信号を出力し、各種の処理の制御を行う。
 <画像処理部の構成、動作>
 次に、画像処理部103の構成と動作について説明する。図3に画像処理部103の構成を示す。画像処理部103は、上記したように、広ダイナミックレンジ画像の生成処理や、動き補正処理などを行う。画像処理部103には、撮像素子102からの補正前画像151が入力される。補正前画像151は、HDR処理部131に供給される。
 HDR処理部131は、供給された補正前画像151から、HDR画像55を生成し、動き補正部132に供給する。補正前画像151は、長時間露光画像と短時間露光画像であり、HDR処理部131は、供給された長時間露光画像と短時間露光画像から、HDR画像55を生成し、動き補正部132に供給する。
 HDR処理部131は、HDR画像55を生成する(合成画像を生成する)ときに、合成統計量152も生成し、動き補正部132に供給する。合計統計量152は、図4を参照して後述するように、HDR画像55を生成するときに、長時間露光画像と短時間露光画像をそれぞれ参照した時間に係わる統計量である。
 また合計統計量152は、後述する例では、ヒストグラムの結果や、そのヒストグラムから求められる参照時刻重心とすることができる。合成統計量152として、ヒストグラムの結果が、動き補正部132に供給され、動き補正部132でヒストグラムから参照時刻重心が求められるようにしても良いし、合成統計量152として、参照時刻重心が、動き補正部132に供給されるようにしても良い。
 動き補正部132は、供給されたHDR画像55に対して、合成統計量152に基づき参照する動き量情報を設定し、参照した動き量情報を基に、動きによる影響を抑制するための補正量を算出し、補正を行う。動き補正部132は、動き補正を行った補正後画像153を、信号処理部104(図2)に出力する。
 <第1の実施の形態:フレーム間での動き補正量の取得>
 HDR処理部131と動き補正部132の処理について、さらに説明を加える。図4は、撮像フレーム、記録フレーム、合成画参照時刻ヒストグラム、動き量情報、および参照範囲の関係を示した図である。
 撮像フレームとしては、長時間露光画像と短時間露光画像がそれぞれあり、その長時間露光画像と短時間露光画像を合成することで、1枚の記録フレームが生成される。例えば、時刻t1から時刻t3が長時間露光期間とされ、この長時間露光期間内に撮像素子102で撮像された画像が、長時間露光画像1Lとして、画像処理部103(内のHDR処理部131)に供給される。
 また、時刻t2から時刻t4が短時間露光期間とされ、この短時間露光期間内に撮像素子102で撮像された画像が、短時間露光画像1Sとして、画像処理部103(内のHDR処理部131)に供給される。そして時刻t1から時刻t4までに得られた長時間露光画像1Lと短時間露光画像1Sから、記録フレーム1F(HDR画像55)が生成される。
 同様に、時刻t4から時刻t7までに得られた長時間露光画像2Lと短時間露光画像2Sから、記録フレーム2F(HDR画像55)が生成され、時刻t7から時刻t10までに得られた長時間露光画像3Lと短時間露光画像3Sから、記録フレーム3F(HDR画像55)が生成され、時刻t10から時刻t13までに得られた長時間露光画像4Lと短時間露光画像4Sから、記録フレーム4F(HDR画像55)が生成される。
 例えば、時刻t1から時刻t4までに得られた長時間露光画像1Lと短時間露光画像1Sから、記録フレーム1F(HDR画像55)が生成されるとき、図1を参照して説明したように、長時間露光画像1Lと短時間露光画像1Sは、ブレンド係数βにより所定の比率で合成される。ブレンド係数βは、例えば、次式により算出される。
 β=(VS)/(VS+VL)
 ただし、
 VL:露出補正後の長時間露光画像の注目画素のノイズの分散値
 VS:露出補正後の短時間露光画像の注目画素のノイズの分散値
 ただし、長時間露光画素が飽和する条件では、β=0となるように補正が行われる。
 ブレンド係数βは、画素毎(フレーム毎、フレーム内のライン毎など設定による)に異なる可能性がある。長時間露光画像と短時間露光画像の合成比率は、画素毎に異なる可能性があり、一定ではない。
 図5は、仮に、長時間露光期間に取得されたデータと短時間露光時間に取得されたデータを、それぞれフレームメモリ内に展開し、その展開された長時間露光画像1Lと短時間露光画像1Sから、記録フレーム1Fを生成した場合の長時間露光画像1L、短時間露光画像1S、記録フレーム1Fの一部(画素配置)を示した図である。
 例えば、長時間露光画像1Lの画素PL1と短時間露光画像1Sの画素PS1から、記録フレーム1Fの画素PF1が生成される。長時間露光画像1Lの画素PL1、短時間露光画像1Sの画素PS1、および記録フレーム1Fの画素PF1は、それぞれの画像内で同一の位置にある画素である。同様に、長時間露光画像2Lの画素PL2と短時間露光画像2Sの画素PS2から、記録フレーム2Fの画素PF2が生成される。
 例えば、長時間露光画像1Lの画素PL1と短時間露光画像1Sの画素PS1から、記録フレーム1Fの画素PF1が生成されたとき、画素PL1と画素PS1のブレンド比は、上記したブレンド係数βにより設定される。画素PL1と画素PS1のブレンド比は、例えば2:1であったとする。
 また例えば、長時間露光画像2Lの画素PL2と短時間露光画像2Sの画素PS2から、記録フレーム2Fの画素PF2が生成されたとき、画素PL2と画素PS2のブレンド比は、例えば3:1であったとする。
 このように、画素毎にブレンド係数βが異なり、ブレンド比が異なる可能性がある。HDR処理部131は、ブレンド係数βを設定し、そのブレンド係数βに基づき、長時間露光画像Lの画素PLと短時間露光画像Sの画素PSを合成することで、HDR画像55(記録フレームF)の画素PFを生成する。またHDR処理部131は、このようにしてHDR画像55を生成するとともに、合成画参照時刻ヒストグラム(図4)も生成する。
 合成画参照時刻ヒストグラム(図4)は、横軸を時刻とし、縦軸が頻度である。頻度は、例えば、ブレンド比(ブレンド係数β)とすることできる。例えば、長時間露光画像1Lの画素PL1と画素PL2が同時刻(時刻tLとする)に取得された画素値であり、短時間露光画像1Sの画素PS1と画素PS2が同時刻(時刻tSとする)に取得された画素値であるとする。また、上記したように、画素PL1と画素PS1のブレンド比は、2:1であり、画素PL2と画素PS2のブレンド比は、3:1であったとする。
 このような場合、時刻tLにおける合成画参照時刻ヒストグラムの値は、5(=2+3)となる。また時刻tSにおける合成画参照時刻ヒストグラムの値は、2(=1+1)となる。このようにして、時刻毎の値(ブレンド比)が算出され、合成画参照時刻ヒストグラムが生成される。なおここでは、ブレンド比(ブレンド係数βから求められる値)を例に挙げて説明したが、他の値を用いてヒストグラムを作成したりすることも可能である。
 このようにして、フレーム毎の合成画参照時刻ヒストグラムが生成される。そして、生成された合成画参照時刻ヒストグラムにおける重心が求められ、その重心の位置が、参照時刻重心とされる。この参照時刻重心の求め方については、後述する。
 この参照時刻重心から、次の参照時刻重心までの間の時間(区間)に取得される動き量情報が用いられて、動き補正量が算出される。例えば、記録フレーム1Fが処理されたときの合成画参照時刻ヒストグラムから時刻T1が、参照時刻重心に設定され、記録フレーム2Fが処理されたときの合成画参照時刻ヒストグラムから時刻T2が、参照時刻重心に設定された場合、時刻T1から時刻T2の区間で得られる動き量情報が、記録フレーム2Fで参照する動き量情報とされ、動き補正量が算出される。
 動き量情報は、例えば、加速度センサやジャイロセンサなどのセンサから、所定の間隔(周期)毎に出力されている。よってこの場合、区間が長くなれば、参照される動き量情報が多くなり、区間が短くなれば、参照される動き量情報は少なくなる。また、参照時刻重心が、長時間露光期間側寄りにある場合には、長時間露光期間内に得られる動き量情報が支配的になり、短時間露光期間側寄りにある場合には、短時間露光期間内に得られる動き量情報が支配的になる。
 図4を参照した説明に戻る。合成画参照時刻ヒストグラムを参照するに、記録フレーム1Fから求められた合成画参照時刻ヒストグラムから、記録フレーム4Fから求められた合成画参照時刻ヒストグラムは、徐々に、長時間露光画像Lから求められるヒストグラムは小さくなり、短時間露光画像Sから求められるヒストグラムは大きくなっている。このような変化にともない、参照時刻重心も、短時間露光画像Sから求められるヒストグラム側に移動している。
 このように、異なるタイミング、例えば、異なる露光時間で撮像された複数の画像を合成するとき、その合成比率に基づき、動きなどによるブレ量を補正するために参照される情報の区間が設定される。よって、より適切な補正を行うことができる。
 ここで、より適切な補正が行えることについてさらに説明を加える。図6は、図4と同じく撮像フレームと記録フレームを表し、上記したようにして参照時刻重心を算出した場合と、従来の方法で参照時刻重心を算出した場合を比較するための図である。
 図6中、C1は、長時間露光期間を表し、C2は、短時間露光期間を表す。また、B2は、長時間露光期間と短時間露光期間の露光間隔を表す。また、図6中、T1’などとダッシュを付した時刻は、従来の方法で算出された参照時刻重心を表し、ダッシュを付していない時刻は、本技術を適用して算出された参照時刻重心を表す。
 図6に示した例では、記録フレームレートを30(ftp)とし、長時間露光期間C1を31(msec)とし、短時間露光期間C2を0.1(msec)とし、露光間隔B2を1(msec)とする。また、ここでは、ブレンド比は、短時間露光画像を全て利用し、長時間露光画像をブレンドしないと設定された場合であるとする。
 このような状況の場合、上記した本実施の形態によれば、短時間露光画像側に参照時刻重心が設定される。図6に、時刻T1、時刻T2、時刻T3、時刻T4として示したように、短時間露光画像側、具体的には、短時間露光期間の中央時刻に参照時刻重心が設定される。
 一方で、このような状況の場合、従来の参照時刻重心の設定によれば、例えば、1記録フレームの中心の時刻、換言すれば、長時間露光期間C1、短時間露光期間C2、および露光間隔B2を加算した中央の時刻が、参照時刻重心に設定される。図6に、時刻T1’、時刻T2’、時刻T3’、時刻T4’として示したように、長時間露光期間の略中央の時刻に参照時刻重心が設定される。
 このような状況の場合、本実施の形態により求められる参照時刻重心と、従来の方式により求められる参照時刻重心とは、大きく異なる場合がある。図6に示した状況では、本実施の形態により求められる参照時刻重心(例えば、時刻T1)と、従来の方式により求められる参照時刻重心(例えば、時刻T1’)との差分は、以下の値になる。
 差分=(C1+B2+C2/2)―(C1+B2+C2)/2
   =(C1+B2)/2
   =(31+1)/2
   =16(msec)
 この場合、記録フレームは、短時間露光画像から生成されているため、短時間露光画像が撮像されたときに得られる動き量情報が支配的になるようにし、動き補正がなされるようにすることで、適切な動き補正が行えると考えられる。本実施の形態により求められる参照時刻重心によれば、短時間露光画像が撮像されたときに動き量情報が支配的になるため、適切な動き補正が行える。
 しかしながら、従来の方式により求められる参照時刻重心は、ブレンド比に係わらず、記録フレームの中央の時刻が設定されるため、図6に示したような状況でも、長時間露光画像が撮像されたときの動き量情報が支配的になり、適切な動き補正が行えない可能性がある。
 図7に、参照時刻重心のずれと、動き補正の効果をシミュレーションした結果を示す。図7に示したグラフにおいて、横軸は、動き周波数(Hz)を表し、縦軸は補正残り量を表す。また図7には、参照時刻重心のずれとして、2,4,8,12,16,20,24,28,32(msec)毎のグラフを示した。
 図中、補正のこり量が100%のラインは、補正の効果がない(0%)であり、100%以上である場合、逆方向に補正してしまう(誤った補正をしてしまう)ことを示している。参照時刻重心のずれが大きくなると、また、動き周波数が大きくなると、補正効果0%のラインを超えやすくなることが読み取れる。
 また、参照時刻重心にずれがあると、手振れ周波数が小さくても、補正残り量が発生し、補正が仕切れていない状況になることが読み取れる。
 このことから、適切な参照時刻重心とのずれがあると、補正のこりが発生し、適切な参照時刻重心とのずれが大きくなると、動き補正の効果が得られづらくなり、さらには、誤補正になってしまう可能性もあることが読み取れる。
 本技術によれば、上記したように、適切な参照時刻重心を設定することができるため(参照時刻重心のずれをほぼ0にすることができるため)、適切な補正を行うことが可能となる。
 <第2の実施の形態:フレーム内での動き補正量の取得>
 第1の実施の形態においては、フレーム間で動き補正量を算出し、動き補正を行う例を挙げて説明した。換言すれば、1フレームにおいて、1個の動き補正量を算出し、そのフレームに対して動き補正を行う実施の形態について説明した。
 第2の実施の形態として、フレーム内で複数の動き補正量を算出し、その算出された動き補正量により、動き補正がなされる例を挙げて説明する。ここでは、フレーム内のライン毎に動き補正量が算出され、ライン毎に動き補正がなされる場合を例に挙げて説明する。
 図8は、図4と同じく、撮像フレーム、記録フレーム、合成画参照時刻ヒストグラム、動き量情報、および参照範囲の関係を示した図である。例えば、撮像フレーム(記録フレーム)の横方向(y軸方向)に線を示したラインyにおいて合成画参照時刻ヒストグラムが生成される。
 第2の実施の形態においては、ライン毎にヒストグラムが生成され、そのヒストグラムに基づいて、ライン毎に参照時刻重心が求められ、動き量情報の参照範囲が設定され、動き補正量が算出され、動き補正が行われる。
 例えば、ラインyは、長時間露光時間内の時刻t1’から時刻t2’の間に露光されていたラインであり、短時間露光時間内の時刻t3’から時刻t4’の間に露光されていたラインである。この場合、長時間露光画像内のラインyに係わるヒストグラムは、時刻t1’から時刻t2’の間に設定されたブレンド比が参照されて作成される。また、短時間露光画像内のラインyに係わるヒストグラムは、時刻t3’から時刻t4’の間に設定されたブレンド比が参照されて作成される。
 このようにして作成されたヒストグラムが用いられ、記録フレーム内のラインyにおける参照時刻重心が設定される。
 このような処理がライン毎に行われる点以外は、フレーム毎に行う第1の実施の形態と同じように処理することが可能なため、その詳細な説明は省略する。
 <参照時刻重心の求め方>
 上記したように、本技術においては、長時間露光画像と短時間露光画像のブレンド比を考慮して、参照時刻重心が求められる。ここで、参照時刻重心の求め方について説明を加える。以下の説明においては、図4における時刻t1乃至t4(長時間露光時間)が16(msec)である場合を例に挙げて説明する。
 参照時刻行列を以下のように定義する。
 T={1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16}  ・・・(1)
ここでは、16(msec)の場合を例示したが、要素数はN個とすることができる。
 式(1)において、各要素は、Vsyncからの経過時間を示す。例えば、1つ目の要素は、Vsyncからの経過時間が0~1msecの区間に対応する。ここでは、ヒストグラムの計算の精度を1msecとしているが、この精度に限定されない。
 次に、画素[y][x]の参照時刻識別行列を以下のように定義する。
 A[y][x]=画素[y][x]の参照時刻識別行列  ・・・(2)
 参照時刻識別行列A[y][x]の一例を以下に示す。例えば、CMOSセンサで一般的なAD変換器の構成であるカラムADを例とすると、x方向の画素については、同一時刻で撮像が行われる。この場合、式(2)の一例は、以下のように示すことができる。
 A[y][x]
 =A[y][0]
 ={{0,0,1,1,1,1,1,1,1,1,1,1,0,0,0,0},{0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1}}
 ここで、1つ目の要素は、画素[y][x]における長時間露光画像の参照時刻識別行列、2つ目の要素は、画素[y][x]における短時間露光画像の参照時刻識別行列である。また、上記、参照時刻識別行列内の数値は、参照している場合に1、参照していない場合に0となる識別情報である。
 次に、合成率行例を以下のように定義する。
 B[y][x]=画素[y][x]の合成率行列  ・・・(3)
 合成率行列B[y][x]の一例を以下に示す。
 B[y][x]={0.8,0.2}
この例の場合、所定の画素において、長時間露光画像を0.8、短時間露光画像を0.2の割合で合成することを意味する。
 同一ライン(y座標)の合成画像参照時刻ヒストグラムhist[y]を以下のように定義する。
Figure JPOXMLDOC01-appb-M000001
 式(4)において、nは、n=0,1,・・・N-1をとる値であり、Nは、式(1)におけるTの要素数を表す。また、mは、参照撮像フレーム番号を表し、長時間露光画像と短時間露光画像の2フレームを参照する場合、m=0,1となる。
 合成画像参照時刻ヒストグラムhist[y]の一例を示す。以下の例においては、B[y][x]は、xによらず{0.8,0.2}であり、同一行で同一合成率である場合を例示している。
 また以下の例は、上記した式(1)の一例、式(2)の一例、式(3)の一例を用いた場合を示している。
 hist[y]={0,0,0.8,0.8,0.8,0.8,0.8,0.8,0.8,0.8,0.8,0.8,0,0.2,0.2,0.2}
 ここまでの処理で、ライン毎にヒストグラムが生成される。第2の実施の形態のように、ライン毎にヒストグラムが生成され、ライン毎に参照時刻重心が算出される場合、式(4)において、ライン毎のヒストグラムが生成され、次式(5)により、ライン毎の参照時刻重心が算出される。
 合成画像参照時刻ヒストグラムの時刻重心center(hist[y])、すなわち参照時刻重心は、次式により求められる。
Figure JPOXMLDOC01-appb-M000002
 合成画像参照時刻ヒストグラムの時刻重心center(hist[y])の一例を示す。上記したhist[y]の一例と同じく、上記した式(1)の一例、式(2)の一例、式(3)の一例を用いた場合であり、式(4)の一例を用いた場合を示している。
 center(hist[y])=(0・1,0・2,0.8・3,0.8・4,0.8・5,0.8・6,0.8・7,0.8・8,0.8・9,0.8・10, 0.8・11,0.8・12,0・13,0.2・14, 0.2・15,0.2・16)/16=4.3125 (msec)
 このようにして、ライン毎に、参照時刻重心が算出される。
 第1の実施の形態のように、フレーム毎にヒストグラムを作成し、フレーム毎に参照時刻重心を算出する場合、以下の式でヒストグラムが作成される。
Figure JPOXMLDOC01-appb-M000003
 式(6)において、hist[y]は、式(4)で算出されるヒストグラムである。式(6)は、式(4)に示したライン毎とのヒストグラム(横方向のヒストグラム)を、縦方向に積算することでフレームのヒストグラムを生成する式である。
 式(6)に基づき算出される1フレームの合成画像参照時刻ヒストグラムから、合成画像参照時刻ヒストグラムの時刻重心center(hist)、すなわち参照時刻重心は、次式により求められる。
Figure JPOXMLDOC01-appb-M000004
 式(7)により、フレーム全体の合成画像参照時刻ヒストグラムの時刻重心、すなわち1フレームにおける参照時刻重心が求められる。
 なお、ここでは、例えば、図4または図8を参照し、ヒストグラムを作成してから、参照時刻重心が求められるとして説明をしたが、ヒストグラムを作成するステップを省略し、直接参照時刻重心が求められるようにしても良い。
 例えば、上記した式(4)(または式(6))において、合成画像参照時刻ヒストグラムを求めるステップを省略し、式(5)(または式(7))で合成画像参照時刻ヒストグラムの時刻重心を算出するようにしても良い。
 <第3の実施の形態:画像処理部の他の構成>
 次に、画像処理部103の他の構成について説明する。図9は、画像処理部103の他の構成を示す図である。図9に示した画像処理部103は、図3に示した画像処理部103と同じく、広ダイナミックレンジ画像の生成処理や、動き補正処理などを行う。
 図3に示した画像処理部103には、撮像素子102からの補正前画像231が入力される。補正前画像231は、動き補正部201に供給される。動き補正部201は、供給された補正前画像231に対して動き補正を行う。動き補正部201には、HDR処理部202から出力された合成統計量234が、フィードバック部203を介して供給され、動き補正部201は、供給された合成統計量234に基づき、補正前画像231に対して動き補正を行う。
 補正前画像231は、上記した例においては、長時間露光画像と短時間露光画像であり、動き補正部201は、長時間露光画像と短時間露光画像に対してそれぞれ動き補正を行う。
 動き補正部201により動き補正が施された補正後画像232は、HDR処理部202に供給される。HDR処理部202は、供給された補正後画像232(例えば、動き補正が施された長時間露光画像と短時間露光画像)から、HDR画像233を生成し、信号処理部104(図2)に供給する。
 HDR処理部202は、HDR画像233を生成する(合成画像を生成する)ときに、合成統計量234も生成し、動き補正部201にフィードバック部203を介して供給する。合計統計量234は、図4などを参照して上述したように、HDR画像233を生成するときに、長時間露光画像と短時間露光画像をそれぞれ参照した時間に係わる統計量(ヒストグラム、またはヒストグラムから求められる参照時刻重心)である。
 図3に示した画像処理部103は、HDR画像を生成した後、動き補正を行う処理フローであるが、図9に示した画像処理部103は動き補正を行った後、HDR画像を生成する処理フローである点が図3に示した画像処理部103と異なる。
 HDR画像を生成するときに、合成統計量が生成され、その合成統計量が用いられて、動き補正が行われる点は、図3および図9にそれぞれ示した画像処理部103において同様であるため、第1の実施の形態および第2の実施の形態として説明した実施の形態を、図9に示した画像処理部103に対しても適用できる。よってここでは、図9に示した画像処理部103の処理についての詳細な説明は省略する。
 図9に示した画像処理部103においては、動き補正を行った後、HDR画像を生成する処理フローであるため、換言すれば、フィードバック部203により合成統計量234がフィードバックされる処理フローであるため、さらに換言すれば、n番目のフレームをHDR画像に変換したときの合成統計量234が用いられて、n+1番目のフレームが生成されるため、1V遅延した合成統計量234で、動き補正が行われることになる。
 しかしながら、合成統計量234は、フレーム毎に大きく変化するような情報ではないため、このように遅延が生じても精度が大きく低下するようなことはない。
 本技術によれば、合成画像が撮像された時刻を基準とした動き補正が可能となり、合成結果によって、動き補正の補正精度が低下するようなことを防ぐことができ、動き補正の効果を100%に近い状態で発揮できる。
 なお、ここでは、長時間露光画像と短時間露光画像を撮像し、HDR画像を生成するときの動き補正を例に挙げて説明したが、本技術は、このような撮像、補正にのみ適用範囲が限定されるものではない。
 本技術は、異なる撮影タイミングで撮影した複数の画を合成する場合に適用できる。複数の画とは、上記した例では、長時間露光画像と短時間露光画像の2枚の画像であるが、2以上の画像が合成されるような場合にも、本技術を適用することはできる。
 例えば、異なる撮影タイミングで撮影した複数の画を合成することで、ノイズ低減の処理を行う場合にも本技術は適用できる。すなわち、本技術は、動き補正の他に、ノイズ低減処理にも適用できる。
 また、上述した実施の形態においては、長時間露光画像が撮像された後、短時間露光画像が撮像される例を挙げて説明した。すなわち、時分割で長時間露光画像と短時間露光画像が撮像される場合を例に挙げて説明した。しかしながら、本技術は、時分割以外、例えば、空間分割で長時間露光画像と短時間露光画像が撮像されるような場合にも適用できる。
 例えば、長時間露光画像を撮影する画素と短時間露光画像を撮像する画素が、画素アレイ部(不図示)に配置され、長時間露光画像が撮影されている間(長時間露光期間中)に、短時間露光画像が撮影される(短時間露光期間がある)場合にも、本技術は適用できる。空間分割で長時間露光画像と短時間露光画像が撮像される場合も、長時間露光画像と短時間露光画像と異なるタイミングで撮像された画像であり、そのような異なるタイミングで撮像された画像を合成するような場合に、本技術は適用できる。
 また、本技術によれば、異なるタイミングで撮像された画像を合成するときの合成比率(ブレンド比)に応じて参照時刻重心が設定され、動き量情報を参照する区間が設定されるため、適切な区間を設定することができる。そのような適切な区間が設定されることで参照される動き量情報を用いた処理としては、上述した動き補正以外の補正(処理)であっても良く、動き補正以外の処理を行うような場合であっても、その処理の精度を向上させることができる。
 また、上述した実施の形態においては、長時間露光画像と短時間露光画像が撮像される場合を例に挙げて説明したが、例えば、IR(赤外線)が投光されることで撮影された画像と、IRが投光されていないときに撮影された画像(環境光で撮影された画像)を合成するような場合にも、本技術を適用することはできる。
 <記録媒体について>
 上述した一連の処理は,ハードウエアにより実行することもできるし,ソフトウエアにより実行することもできる。一連の処理をソフトウエアにより実行する場合には、そのソフトウエアを構成するプログラムが、コンピュータにインストールされる。ここで、コンピュータには、専用のハードウエアに組み込まれているコンピュータや、各種のプログラムをインストールすることで、各種の機能を実行することが可能な、例えば汎用のパーソナルコンピュータなどが含まれる。
 図10は、上述した一連の処理をプログラムにより実行するコンピュータのハードウエアの構成例を示すブロック図である。コンピュータにおいて、CPU(Central Processing Unit)501、ROM(Read Only Memory)502、RAM(Random Access Memory)503は、バス504により相互に接続されている。バス504には、さらに、入出力インタフェース505が接続されている。入出力インタフェース505には、入力部506、出力部507、記憶部508、通信部509、およびドライブ510が接続されている。
 入力部506は、キーボード、マウス、マイクロフォンなどよりなる。出力部507は、ディスプレイ、スピーカなどよりなる。記憶部508は、ハードディスクや不揮発性のメモリなどよりなる。通信部509は、ネットワークインタフェースなどよりなる。ドライブ510は、磁気ディスク、光ディスク、光磁気ディスク、又は半導体メモリなどのリムーバブルメディア511を駆動する。
 以上のように構成されるコンピュータでは、CPU501が、例えば、記憶部508に記憶されているプログラムを、入出力インタフェース505およびバス504を介して、RAM503にロードして実行することにより、上述した一連の処理が行われる。
 コンピュータ(CPU501)が実行するプログラムは、例えば、パッケージメディア等としてのリムーバブルメディア511に記録して提供することができる。また、プログラムは、ローカルエリアネットワーク、インターネット、デジタル衛星放送といった、有線または無線の伝送媒体を介して提供することができる。
 コンピュータでは、プログラムは、リムーバブルメディア511をドライブ510に装着することにより、入出力インタフェース505を介して、記憶部508にインストールすることができる。また、プログラムは、有線または無線の伝送媒体を介して、通信部509で受信し、記憶部508にインストールすることができる。その他、プログラムは、ROM502や記憶部508に、あらかじめインストールしておくことができる。
 なお、コンピュータが実行するプログラムは、本明細書で説明する順序に沿って時系列に処理が行われるプログラムであっても良いし、並列に、あるいは呼び出しが行われたとき等の必要なタイミングで処理が行われるプログラムであっても良い。
 また、本明細書において、システムとは、複数の装置により構成される装置全体を表すものである。
 なお、本明細書に記載された効果はあくまで例示であって限定されるものでは無く、また他の効果があってもよい。
 なお、本技術の実施の形態は、上述した実施の形態に限定されるものではなく、本技術の要旨を逸脱しない範囲において種々の変更が可能である。
 なお,本技術は以下のような構成も取ることができる。
(1)
 異なるタイミングで撮像された複数の画像を合成する合成部と、
 動きに関する前記画像への影響を低減する補正を行う動き補正部と
 を備え、
 前記動き補正部は、
 所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
 前記設定された基準を元に、前記動き量情報を参照し、前記補正を行う
 画像処理装置。
(2)
 前記複数の画像は、長時間露光期間で撮像された長時間露光画像と、短時間露光期間で撮像された短時間露光画像であり、
 前記合成部は、前記長時間露光画像と前記短時間露光画像を前記合成比率で合成することで、ダイナミックレンジの広い画像を生成する
 前記(1)に記載の画像処理装置。
(3)
 前記複数の画像は、同一露光時間で撮像され、異なるタイミングで撮像された複数の画像であり、
 前記合成部は、前記複数の画像を合成することで、ノイズが低減された画像を生成する
 前記(1)に記載の画像処理装置。
(4)
 前記動き補正部は、
 前記合成部により複数の画像が合成されることで生成された合成画像が、前記複数の画像を参照する時刻に関するヒストグラムを統計量とし、
 前記統計量をもとに、前記基準を設定する
 前記(1)乃至(3)のいずれかに記載の画像処理装置。
(5)
 前記基準は、前記合成画像毎に設定される
 前記(4)に記載の画像処理装置。
(6)
 前記基準は、前記合成画像のライン毎に設定される
 前記(4)に記載の画像処理装置。
(7)
 前記合成部により複数の画像が合成されることで生成された合成画像に対して、前記動き補正部による補正が行われる
 前記(1)乃至(6)のいずれかに記載の画像処理装置。
(8)
 前記動き補正部による補正が施された前記複数の画像を、前記合成部は合成する
 前記(1)乃至(6)のいずれかに記載の画像処理装置。
(9)
 異なるタイミングで撮像された複数の画像を合成し、
 動きに関する前記画像への影響を低減する動き補正を行う
 ステップを含み、
 前記動き補正を、
 所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
 前記設定された基準を元に、前記動き量情報を参照して行う
 処理を含む画像処理方法。
(10)
 異なるタイミングで撮像された複数の画像を合成する合成部と、
 動きに関する前記画像への影響を低減する補正を行う動き補正部と
 を備え、
 前記動き補正部は、
 所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
 前記設定された基準を元に、前記動き量情報を参照し、前記補正を行う
 画像処理装置を備える
 撮像装置。
(11)
 コンピュータに、
 異なるタイミングで撮像された複数の画像を合成し、
 動きに関する前記画像への影響を低減する動き補正を行う
 ステップを含み、
 前記動き補正を、
 所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
 前記設定された基準を元に、前記動き量情報を参照して行う
 処理を実行させるためのプログラム。
 100 撮像装置, 101 光学レンズ, 102 撮像素子, 103 画像処理部, 104 信号処理部, 105 制御部, 131 HDR処理部, 132 動き補正部, 201 動き補正部, 202 HDR処理部, 203 フィードバック部

Claims (11)

  1.  異なるタイミングで撮像された複数の画像を合成する合成部と、
     動きに関する前記画像への影響を低減する補正を行う動き補正部と
     を備え、
     前記動き補正部は、
     所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
     前記設定された基準を元に、前記動き量情報を参照し、前記補正を行う
     画像処理装置。
  2.  前記複数の画像は、長時間露光期間で撮像された長時間露光画像と、短時間露光期間で撮像された短時間露光画像であり、
     前記合成部は、前記長時間露光画像と前記短時間露光画像を前記合成比率で合成することで、ダイナミックレンジの広い画像を生成する
     請求項1に記載の画像処理装置。
  3.  前記複数の画像は、同一露光時間で撮像され、異なるタイミングで撮像された複数の画像であり、
     前記合成部は、前記複数の画像を合成することで、ノイズが低減された画像を生成する
     請求項1に記載の画像処理装置。
  4.  前記動き補正部は、
     前記合成部により複数の画像が合成されることで生成された合成画像が、前記複数の画像を参照する時刻に関するヒストグラムを統計量とし、
     前記統計量をもとに、前記基準を設定する
     請求項1に記載の画像処理装置。
  5.  前記基準は、前記合成画像毎に設定される
     請求項4に記載の画像処理装置。
  6.  前記基準は、前記合成画像のライン毎に設定される
     請求項4に記載の画像処理装置。
  7.  前記合成部により複数の画像が合成されることで生成された合成画像に対して、前記動き補正部による補正が行われる
     請求項1に記載の画像処理装置。
  8.  前記動き補正部による補正が施された前記複数の画像を、前記合成部は合成する
     請求項1に記載の画像処理装置。
  9.  異なるタイミングで撮像された複数の画像を合成し、
     動きに関する前記画像への影響を低減する動き補正を行う
     ステップを含み、
     前記動き補正を、
     所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
     前記設定された基準を元に、前記動き量情報を参照して行う
     処理を含む画像処理方法。
  10.  異なるタイミングで撮像された複数の画像を合成する合成部と、
     動きに関する前記画像への影響を低減する補正を行う動き補正部と
     を備え、
     前記動き補正部は、
     所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
     前記設定された基準を元に、前記動き量情報を参照し、前記補正を行う
     画像処理装置を備える
     撮像装置。
  11.  コンピュータに、
     異なるタイミングで撮像された複数の画像を合成し、
     動きに関する前記画像への影響を低減する動き補正を行う
     ステップを含み、
     前記動き補正を、
     所定の間隔で供給される動き量に関する動き量情報を参照する区間の基準を、前記複数の画像を合成するときの合成比率に基づき設定し、
     前記設定された基準を元に、前記動き量情報を参照して行う
     処理を実行させるためのプログラム。
PCT/JP2017/017218 2016-05-16 2017-05-02 画像処理装置、画像処理方法、撮像装置、プログラム Ceased WO2017199758A1 (ja)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US16/087,791 US11064129B2 (en) 2016-05-16 2017-05-02 Image processing apparatus, image processing method, imaging apparatus, and program
CN201780025394.8A CN109076170B (zh) 2016-05-16 2017-05-02 图像处理设备、图像处理方法、成像设备和存储介质

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2016097883A JP2017208604A (ja) 2016-05-16 2016-05-16 画像処理装置、画像処理方法、撮像装置、プログラム
JP2016-097883 2016-05-16

Publications (1)

Publication Number Publication Date
WO2017199758A1 true WO2017199758A1 (ja) 2017-11-23

Family

ID=60326312

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2017/017218 Ceased WO2017199758A1 (ja) 2016-05-16 2017-05-02 画像処理装置、画像処理方法、撮像装置、プログラム

Country Status (4)

Country Link
US (1) US11064129B2 (ja)
JP (1) JP2017208604A (ja)
CN (1) CN109076170B (ja)
WO (1) WO2017199758A1 (ja)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021101037A1 (en) * 2019-11-19 2021-05-27 Samsung Electronics Co., Ltd. System and method for dynamic selection of reference image frame
US10911691B1 (en) * 2019-11-19 2021-02-02 Samsung Electronics Co., Ltd. System and method for dynamic selection of reference image frame
US12063452B2 (en) 2020-01-30 2024-08-13 Sony Group Corporation Control device, imaging system, and 3D model data generation method
KR20210099865A (ko) * 2020-02-05 2021-08-13 삼성전자주식회사 이미지 센서 모듈, 그것의 영상 압축 방법 및 전자 장치
JP2022099651A (ja) * 2020-12-23 2022-07-05 ソニーセミコンダクタソリューションズ株式会社 画像生成装置、画像生成方法、およびプログラム
CN114531552B (zh) * 2022-02-16 2023-06-27 四川创安微电子有限公司 一种高动态范围图像合成方法及系统
CN119865706B (zh) * 2023-10-19 2026-01-06 北京小米移动软件有限公司 拍照控制方法、装置及存储介质

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007202098A (ja) * 2005-12-27 2007-08-09 Kyocera Corp 撮像装置および撮像方法
JP2013066142A (ja) * 2011-08-31 2013-04-11 Sony Corp 画像処理装置、および画像処理方法、並びにプログラム

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4403635B2 (ja) 2000-05-31 2010-01-27 パナソニック株式会社 ダイナミックレンジ拡大カメラ
WO2005079062A1 (ja) * 2004-02-13 2005-08-25 Sony Corporation 画像処理装置と画像処理方法およびプログラム
JP5458865B2 (ja) * 2009-09-18 2014-04-02 ソニー株式会社 画像処理装置、撮像装置、および画像処理方法、並びにプログラム
JP2012010104A (ja) * 2010-06-24 2012-01-12 Sony Corp 画像処理装置、撮像装置、画像処理方法、およびプログラム
JP2013038504A (ja) * 2011-08-04 2013-02-21 Sony Corp 撮像装置、および画像処理方法、並びにプログラム
KR101665513B1 (ko) * 2015-01-30 2016-10-12 삼성전자 주식회사 컴퓨터 단층 촬영 장치 및 그에 따른 ct 영상 복원 방법
KR101725891B1 (ko) * 2015-08-05 2017-04-11 삼성전자주식회사 단층 영상 장치 및 그에 따른 단층 영상 복원 방법

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007202098A (ja) * 2005-12-27 2007-08-09 Kyocera Corp 撮像装置および撮像方法
JP2013066142A (ja) * 2011-08-31 2013-04-11 Sony Corp 画像処理装置、および画像処理方法、並びにプログラム

Also Published As

Publication number Publication date
CN109076170B (zh) 2020-10-27
US11064129B2 (en) 2021-07-13
CN109076170A (zh) 2018-12-21
JP2017208604A (ja) 2017-11-24
US20200221007A1 (en) 2020-07-09

Similar Documents

Publication Publication Date Title
WO2017199758A1 (ja) 画像処理装置、画像処理方法、撮像装置、プログラム
CN105659583B (zh) 图像处理装置、图像处理方法、电子设备及计算机可读存储介质
JP6434963B2 (ja) 画像処理装置、画像処理方法、電子機器、並びにプログラム
US9420193B2 (en) Image pickup apparatus, method of controlling image pickup apparatus, and program
US20120062694A1 (en) Imaging apparatus, imaging method, and program
US20120273655A1 (en) Image sensing apparatus and control method thereof
CN103748868A (zh) 成像设备、信号处理方法及程序
JP2009152669A (ja) 撮像装置、撮像処理方法及び撮像制御プログラム
JP6703847B2 (ja) 撮像装置及び撮像方法
US9667882B2 (en) Image processing apparatus, image-pickup apparatus, image processing method, non-transitory computer-readable storage medium for generating synthesized image data
CN105100601A (zh) 摄像装置及其控制方法和信号处理装置
WO2016129405A1 (ja) 画像処理装置および画像処理方法、並びにプログラム
US20080316333A1 (en) Imaging apparatus, imaging method, program, and integrated circuit
WO2016006430A1 (ja) 撮像装置、撮像方法、およびプログラム、並びに再生装置
TWI692965B (zh) 影像處理方法、電子裝置及非暫態電腦可讀取記錄媒體
JP2020108022A (ja) 撮像装置、撮像システム、移動体及び撮像装置の制御方法
JP2017183870A (ja) 撮像処理装置、撮像処理方法、コンピュータプログラム及び電子機器
JP4739998B2 (ja) 撮像装置
JP2017085337A (ja) 画像処理装置、撮像装置および画像処理プログラム
JP6742823B2 (ja) 画像処理装置、画像処理方法、及びプログラム
JP6141129B2 (ja) 撮像装置、撮像装置の制御方法、及びコンピュータプログラム
JP6706850B2 (ja) 撮像装置、カメラ、及び撮像方法
JP2023055302A (ja) 画像処理装置、撮像装置および画像処理方法
JP4356604B2 (ja) 映像信号処理装置、映像信号の処理方法、映像信号の処理プログラム及び映像信号の処理プログラムを記録した記録媒体
JP2015023314A (ja) 撮像装置、撮像装置の制御方法、及びコンピュータプログラム

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

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

Ref document number: 17799188

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 17799188

Country of ref document: EP

Kind code of ref document: A1