WO2012058977A1 - 多光谱成像颜色测量系统及其成像信号处理方法 - Google Patents

多光谱成像颜色测量系统及其成像信号处理方法 Download PDF

Info

Publication number
WO2012058977A1
WO2012058977A1 PCT/CN2011/078862 CN2011078862W WO2012058977A1 WO 2012058977 A1 WO2012058977 A1 WO 2012058977A1 CN 2011078862 W CN2011078862 W CN 2011078862W WO 2012058977 A1 WO2012058977 A1 WO 2012058977A1
Authority
WO
WIPO (PCT)
Prior art keywords
imaging
image
channel
color measurement
filter wheel
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/CN2011/078862
Other languages
English (en)
French (fr)
Inventor
忻浩忠
邵思杰
沈会良
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.)
Hong Kong Research Institute of Textiles and Apparel Ltd
Original Assignee
Hong Kong Research Institute of Textiles and Apparel 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 Hong Kong Research Institute of Textiles and Apparel Ltd filed Critical Hong Kong Research Institute of Textiles and Apparel Ltd
Priority to US13/882,508 priority Critical patent/US9417132B2/en
Priority to EP11837503.9A priority patent/EP2637004B1/en
Publication of WO2012058977A1 publication Critical patent/WO2012058977A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/46Measurement of colour; Colour measuring devices, e.g. colorimeters
    • G01J3/52Measurement of colour; Colour measuring devices, e.g. colorimeters using colour charts
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02Details
    • G01J3/0205Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows
    • G01J3/0208Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows using focussing or collimating elements, e.g. lenses or mirrors; performing aberration correction
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02Details
    • G01J3/0205Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows
    • G01J3/0216Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows using light concentrators or collectors or condensers
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02Details
    • G01J3/0205Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows
    • G01J3/0235Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows using means for replacing an element by another, for replacing a filter or a grating
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02Details
    • G01J3/10Arrangements of light sources specially adapted for spectrometry or colorimetry
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28Investigating the spectrum
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/46Measurement of colour; Colour measuring devices, e.g. colorimeters
    • G01J3/50Measurement of colour; Colour measuring devices, e.g. colorimeters using electric radiation detectors
    • G01J3/51Measurement of colour; Colour measuring devices, e.g. colorimeters using electric radiation detectors using colour filters
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/46Measurement of colour; Colour measuring devices, e.g. colorimeters
    • G01J3/52Measurement of colour; Colour measuring devices, e.g. colorimeters using colour charts
    • G01J3/524Calibration of colorimeters
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/80Camera processing pipelines; Components thereof
    • H04N23/84Camera processing pipelines; Components thereof for processing colour signals
    • H04N23/88Camera processing pipelines; Components thereof for processing colour signals for colour balance, e.g. white-balance circuits or colour temperature control
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/12Generating the spectrum; Monochromators
    • G01J2003/1226Interference filters
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28Investigating the spectrum
    • G01J3/2823Imaging spectrometer
    • G01J2003/2826Multispectral imaging, e.g. filter imaging
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N2021/1765Method using an image detector and processing of image signal
    • G01N2021/177Detector of the video camera type
    • G01N2021/1776Colour camera

Definitions

  • the present invention relates to the field of imaging, and more particularly to a multi-spectral imaging color measurement system and an imaging signal processing method thereof. Background technique
  • spectrophotometers are available for color measurement of large monochrome samples with high color accuracy.
  • Conventional imaging devices include digital cameras and color scanners that record color information by reflecting light from a surface of the object through a plurality of color filters having different spectral transmittances, and then forming the reflected light into an image using an electronic sensor. They have many advantages in measuring many colors at the same time, especially when measuring small, multi-colored objects and capturing color information for objects with complex textured surfaces.
  • traditional three primary color and chromatic imaging systems are difficult to produce high-color-accurate images because they violate the Luther rule.
  • the Luther rule means that the sensitivity of most cameras is not similar to or similar to the human eye. Not a linear combination of human eyes.
  • this colorimetric imaging system can only output three stimulus values, which can lead to metameric errors caused by different observers and different illumination systems. Therefore, these colorimetric measurement systems, such as Dig iEye manufactured by Verivide Ltd., are not accurate enough for the textile and garment industries with high color quality requirements. Summary of the invention
  • the technical problem to be solved by the present invention is to provide a high-accuracy multi-spectral imaging color measurement system and a multi-spectral imaging color measurement system in the prior art which is insufficiently accurate in the textile and garment industry. Imaging signal processing method.
  • a multi-spectral imaging color measurement system comprising a black chamber for forming a closed space for multi-spectral imaging, a sample stage for placing and fixing a sample to be tested in a black chamber, and a black chamber
  • An imaging system for taking images of a sample to be tested including a controllable illumination system, a filter wheel system, an imaging signal processing system, and an electronic control system.
  • the controllable illumination system is located in the black chamber, and includes at least one lamp tube symmetrically arranged on the side of the sample to be tested and directed to the sample to be tested;
  • the filter wheel system is located between the imaging system and the sample to be tested for filtering light emitted by the controllable illumination system reflected by the sample to be tested;
  • the imaging signal processing system is located within the imaging system for performing calibration and reflection reconstruction of images captured by the imaging system;
  • the electronic control system is in communication with the controllable illumination system, the filter wheel system, and the imaging system for controlling the operational status of each of the above systems.
  • the lamp tube includes a light source sequentially connected, an optical integration column for collecting more uniform light, a lens group for further improving illumination uniformity and magnification, and a photoresist plate for reducing diffused light, the interior of the lamp tube being further coated with a light absorbing material for reducing internal diffused light;
  • the light source is a surface tungsten lamp having a smooth curve spectral energy distribution, and is powered by two high-precision DC stabilized power supplies;
  • the optical integration column is a hollow horn structure surrounded by a glass wall
  • the lens group includes a grating for limiting an edge of the beam, and further includes one or more convex lenses and concave lenses having different refractive indices;
  • the photoresist plate is placed on the front edge of the lens group.
  • the filter wheel system includes a motor, a filter wheel, and a belt device connecting the motor and the filter wheel;
  • the motor is configured to provide the filter wheel with power and communication for filter position selection, controlled by the electronic control system;
  • the filter wheel further includes a chassis having one or more slots and one or more holes for reducing the rotational load, the filter being fixed in the slot by a top buckle
  • the filter wheel further includes a built-in infrared optical switch for position detection;
  • the belt device further includes a first synchronous wheel mounted on a rotating shaft of the motor, mounted in the same a second synchronizing wheel on the shaft of the chassis and a belt that meshes with the outer edges of the first synchronizing wheel and the second synchronizing wheel.
  • the imaging system further includes a CCD sensor or a CMOS sensor, the sensor having an A/D converter built therein for passing the projection onto the focal plane of the sensor
  • the optical signal of the selected wavelength of the filter wheel system is converted into a digital electrical signal to produce a multi-channel spectral image.
  • the electronic control system includes: a lighting system control unit for adjusting voltage and current setting values of the lighting system power source to conform to a stable working state of the light source;
  • a microprocessor unit for controlling three operating states of acceleration, steady state driving, and deceleration of the motor
  • An interface circuit board for communication between the imaging system, the lighting system control unit, and the filter wheel system.
  • the present invention also provides an imaging signal processing method comprising geometrically correcting an error caused by an optical lens or a filter, and performing brightness correction on each optical channel image, and further comprising: estimating an exposure time of the linear working range of the sensor, The sensor operates within the exposure time to convert an incident optical signal into a digital imaging signal; corrects image noise caused by an inherent noise source in the sensor; performs images of different optical channels acquired under different filters Multi-channel image registration to eliminate differences between the different channel images due to the relative positional differences of the filters in the filter wheel, the different refractive index effects of the filters, the lens dispersion, or the small gap between the imaging system and the object being measured Content offset; correcting the image of the overlay caused by the ghosting effect on each channel;
  • the reflectance is reconstructed by the above-described processed imaging signal to generate a spectral reflectance image of the sample to be tested.
  • the intrinsic noise source includes a dark current
  • Ic is used.
  • Rr stands for the corrected image
  • I rep stands for the original or original uncorrected image
  • I Dark stands for the "dark current” image
  • I whits stands for the uniform white target image
  • the coefficient k is a guarantee that the CCD sensor operates in the linear range.
  • the target channel image is subjected to recovery calibration according to the generated mapping function f (X).
  • "correcting the image of the superimposed image caused by the ghosting effect on each channel” further includes:
  • the ghost image for each imaging channel eliminates the ghosting effect according to its associated parameters.
  • "reconstructing the reflectance using the imaged signal after the above processing” specifically includes reconstructing the reflectance using Wiener estimation or pseudo-inverse calibration.
  • the multi-spectral imaging color measurement system and its imaging signal processing method of the present invention have the beneficial effects that the multi-spectral imaging color measurement system can overcome the shortcomings of the conventional digital imaging system and the spectrophotometer system.
  • the complete multi-spectral imaging color measurement (IC ⁇ ) system provides the basics for color measurement and assessment for textile industry users.
  • the present invention designs a uniform illumination system to provide a uniform imaging environment and designs an imaging signal processing system to simultaneously capture high spatial resolution spectral colors. Image. Therefore, the present invention can be easily measured in any number of colors, no matter how small the color is. This technique is superior to spectrophotometers in providing structural information of objects and is superior to current digital camera systems in obtaining true spectral reflectance and eliminating metamerism under different light sources.
  • FIG. 1 is a schematic diagram of an imaging color measurement system in accordance with one embodiment of the present invention.
  • FIG. 2 is a schematic view of a lamp tube in accordance with one embodiment of the present invention.
  • FIG. 3 is a schematic view of an optical integration column in accordance with one embodiment of the present invention.
  • FIG. 4 is a schematic view of a lens system in accordance with one embodiment of the present invention.
  • Figure 5 is a schematic illustration of a symmetric illumination system including two lamps in accordance with one embodiment of the present invention
  • Figure 6 is an annular illumination consisting of eight light sources uniformly distributed at 45° in accordance with another embodiment of the present invention. Schematic diagram of the system;
  • Figure 7A is a schematic illustration of a multi-story illumination system in accordance with another embodiment of the present invention
  • Figure 7B is a schematic illustration of the included angle of a multi-story illumination system in accordance with another embodiment of the present invention
  • FIG. 9 is a flow chart of an imaging signal processing method in accordance with the present invention.
  • Figure 10 is a graph showing the relationship between the average response value of the captured image in each channel and the respective exposure time intervals
  • Figure 11A is a reference channel image during multi-channel image calibration
  • Figure 11B is a target channel image in a multi-channel image calibration process
  • 11C is a schematic diagram showing the positional deviation of the reference channel image and the target channel image before multi-channel image calibration
  • Figure 11D is a schematic diagram showing the positional deviation of the reference channel image and the target channel image after multi-channel image calibration
  • Figure 12 is a schematic illustration of a non-parallel surface of a filter in an imaged color measurement system in accordance with one embodiment of the present invention.
  • Figure 13 is a schematic diagram of a ghosting effect in accordance with one embodiment of the present invention
  • 14 is a schematic diagram of a template matching process in accordance with one embodiment of the present invention.
  • FIG. 1 is a schematic illustration of an imaging color measurement system in accordance with one embodiment of the present invention.
  • all imaging components are housed in the black chamber 6, placing the whiteboard 5 and the sample in two symmetrical tilts at a geometric angle of 45 ⁇ 5°.
  • the lamp tube 4 is under.
  • the controllable lighting system is specially designed.
  • the imaging system 1, the beam splitter wheel 3 and the lens 2 are placed in a top box for multispectral imaging of the object.
  • a narrowband interference filter is used in the multispectral imaging color measurement system of the present invention.
  • each tube there are 3 units, such as a light source, an optical integration column, and a lens system.
  • the surface tungsten lamp is used as a light source to emit continuous light with a stable spectral power distributed in the visible light length and stable in a certain period of time.
  • the optical integrator column is used to concentrate as much light as possible, which produces a uniformly diffused reflected light at the exit of the optical integrator column, which in turn passes through the lens group to further improve uniformity and magnification in the effective range of uniform illumination.
  • a uniform illumination area that meets the requirements of a multispectral imaging color measurement system can be provided.
  • High-pressure xenon lamps can also emit a continuous spectrum, and its color temperature is generally 5000k. If a conversion filter is combined with a xenon lamp, the D65 standard illumination body can be better simulated. But like all other gas discharge sources, the xenon lamp has a specific spectrum of emission spectrum, which is about 475 nm. In order to reduce the influence of the characteristic line, a special beam splitter is usually used to suppress the emission line spectrum. In contrast, the corresponding spectral energy distribution of a tungsten halogen lamp is a smooth curve with no spikes or small jitter. In the present invention, a tungsten halogen lamp is used as a light source to provide a continuous spectral distribution of light for a multispectral imaging color measurement system.
  • the stability of planar light sources is the first step in achieving high quality images.
  • Two high-precision DC regulated power supplies provide stable voltage and current for the surface tungsten lamp to ensure the stability of the light source.
  • a wired or wireless interface is used to connect the DC regulated power supply to the host. The host sends a command to adjust the required voltage and current values to meet the set value of the stable operation of the light source according to some specified wired or wireless protocols.
  • each of the lamps further includes a photoresist plate.
  • a photoresist plate is placed on the front edge of each lens group to reduce the diffusion of the surface of the illuminated object.
  • Light In order to further reduce the diffused light, the entire light tube can also be coated with a light absorbing material.
  • 3 is a schematic illustration of an optical integration column in accordance with one embodiment of the present invention. As shown in FIG. 3, the multi-reflected emitted light in the tungsten halogen lamp is regarded as a point source having a certain scale and high brightness. The brightness distribution of this point source is similar to the Gaussian distribution, which is not uniform.
  • the emitted light is further processed.
  • an important optical device is designed to guide light, i.e., an optical integrating column. Its main role is to collect more uniform light.
  • the integrating column is a hollow structure surrounded by an optical glass sheet for homogenizing the emitted light.
  • This horn structure is used to collect as much light as possible, for example to obtain a diffusely reflected light from the left side of the optical integration column.
  • the angle of divergence of the beam introduced into the integrating column is smaller than the angle of the aperture of the light guide, light that is reflected multiple times through the inner wall of the glass will be emitted from the exit of the integrating column.
  • the spread angle of the beam is greater than the aperture angle of the light guide, the beam will refract. Through this light guiding mechanism, the light coming out from the other end of the integrating column will form a more uniform intensity beam, and the direction of light propagation is messy.
  • the aperture angle of the photorefractive tube is determined by its refractive index.
  • FIG. 4 is a schematic illustration of a lens system in accordance with one embodiment of the present invention.
  • the light that is specularly reflected from the optical integrator column can be regarded as uniform, the effective range of the uniform illumination region of the light is relatively small, and cannot satisfy the large scale on the measurement plane in the multispectral imaging color measurement system. The requirement for a uniform lighting area.
  • a series of lenses are used to expand the uniform illumination area of the light emerging from the optical integrator.
  • the lens system used in the present invention includes a plurality of lenses, and the specific optical path diagram is as shown in FIG.
  • a lens system employed in one embodiment of the invention includes two lens groups.
  • the first group consists of five lenses, two of which are concave lenses and three convex lenses.
  • the second group includes a concave lens and a convex lens.
  • the combination of a concave lens and a convex lens is used to eliminate the geometric distortion of the lens group.
  • Lenses with different refractive indices are selected in the lens system to eliminate dispersion errors.
  • the primary role of the grating in the lens system is to limit the beam edges from being projected to the outside. In general, the distortion of the beam edge is even more severe than the distortion caused by changing the lens position.
  • the grating can effectively increase the brightness and contrast of the screen brightness, but at the same time limit the utilization of the light source.
  • Figure 5 is a schematic illustration of a symmetrical illumination system with two lamps, in accordance with one embodiment of the present invention. As shown in Figure 5, the two tubes are symmetrically placed at a geometric angle of 45 ⁇ 5°. Although the illumination system shown in Figures 1 and 5 employs two symmetrical lamps, it is not limited to two symmetrical tubes. In another embodiment of the invention, the ring illumination system of the present invention may Includes any number or combination of tungsten halogen lamp sources. Additionally, the illumination system of the present invention may also include any number of layers of light sources to create a uniform illumination area.
  • Figure 6 shows a schematic diagram of a ring-shaped symmetrical illumination system in accordance with another embodiment of the present invention. As shown in Fig. 6, the eight light sources are evenly distributed on the ring at intervals of 45°.
  • Figures 7A and 7B further illustrate schematic views of a multilayer annular illumination system.
  • Each of the light sources of each layer in Figure 7 is represented by a circle.
  • Each layer of this multilayer annular illumination system can work independently.
  • a multi-layered annular illumination system can illuminate the measured object from different angles to obtain a more accurate spectral analysis of the directional texture pattern sample and other objects whose colors depend on different incident angle effects of the light source.
  • FIG. 8 is a schematic illustration of a filter wheel system in accordance with one embodiment of the present invention.
  • the filter wheel includes a plurality of narrow band filters mounted on the teeth, and the narrow band filters are continuously and closely arranged on the wheel.
  • These wavelength-selectable narrow-band filters allow light of different wavelength ranges to pass through the corresponding filters. For example, different wavelengths of these narrowband filters can be used to filter light into different spectra of 10 or 20 nanometers wide.
  • different regions of the wavelength selective narrow band pass filter can filter light to 400 nm, 420 nm, 440 nm, 460 nm, 480 nm, 500 nm, 520 nm, 540 nm, 560 nm, 580 nm, 600 jin, 620 jin, 640 jin 660, 680, and 700 are centered on different 20-inch bandwidth spectra.
  • the number of narrow-band filters is not limited to 16 or 31, and the bandwidth of the filters is not limited to 20 or 10 nm.
  • Figure 8 illustrates an embodiment of the invention in which the filter wheel includes 16 narrow band filters.
  • a custom filter wheel with a wavelength-selectable narrow-band filter can be used between the imaging system and the object to be measured to acquire information for each spectral channel, making the imaging system a color analyzer. /sensor.
  • the narrow band filter provides a spectrum of the appropriate wavelength range for spectral analysis and color determination.
  • a wavelength-selectable narrow-band filter can provide spectra in 16 different wavelength ranges for detection and analysis, and of course, other suitable numbers of filters or spectra.
  • the filter wheel used in this embodiment includes a chassis having 16 slots for arranging filters. In each slot, use a top snap ring to secure the edges of the filter and reduce the filter rotation The tilt of the filter when turning.
  • a belt device is used instead of the direct coupling to the stepper motor drive shaft to carry a heavier load and ensure higher rotational accuracy.
  • the first synchronous wheel is mounted on the rotating shaft of the stepping motor, and the second synchronous wheel is mounted on the periphery of the wheel chassis shaft.
  • a plurality of narrow-band filters of equidistant annular arrangement are arranged between the central axis and the periphery of the wheel chassis. .
  • a variety of holes in the wheel chassis are used to reduce the load when rotating.
  • the belt that meshes with the outer edges of the first synchronizing wheel and the second synchronizing wheel functions to drive the filter wheel chassis to rotate centering on the first synchronizing wheel.
  • the filter wheel shown in Figure 8 is a compact disc design with a stepper motor attached to one end.
  • the filter wheel has a built-in infrared optical switch for position detection that allows initial position calibration of the wheel at the beginning of each rotation.
  • the stepper motor is controlled by a microprocessor module that connects the filter wheel to the main unit via a wired or wireless interface.
  • the stepper motor is an external component that provides the motorized filter wheel with power and communication for filter position selection.
  • two complementary controls are designed to detect the rotational speed and position of the filter wheel mechanism. Automatic control is achieved using an incremental encoder that gives a reference position.
  • the stepper motor is normally energized even when it is stopped, and there is no significant bounce or positional drift when it is fixed in the slot.
  • the microprocessor sends a rotation pulse of three working states to the driver of the motor.
  • the three operating states include acceleration, steady-state driving and deceleration, and at the same time, the starting frequency, the driving frequency, the acceleration time, and the deceleration.
  • the time is set to the corresponding predetermined value.
  • the beam propagates through the filter and optical lens at a selected wavelength. Then, it is projected onto the focal plane of a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS) sensor.
  • CCD or CMOS sensor can convert the analog optical signal into digital power through an A/D converter built into the imaging sensor circuit. signal.
  • the stepper motor drives the wheel to rotate continuously, and the imaging sensor produces a spectral image of 16 or more channels.
  • an interface board is installed.
  • the interface can support any suitable type of communication medium, such as a wired or wireless network or connection.
  • the interface can include any structure suitable for communication between the microprocessor and the host. For example, it may be an interface board that supports the synchronization mechanism of the camera exposure time, or an interface board that supports continuous positioning of narrow-band filters connected in accordance with the RS232 communication protocol.
  • Figure 9 is a flow chart of an imaging signal processing method in accordance with the present invention.
  • the original image Imaging signal processing methods like conversion to photometric signals include: exposure time estimation, imaging noise correction, multi-channel image calibration, ghosting correction, image brightness correction, and reflectance reconstruction.
  • the geometric correction and brightness correction of the moving lens are technical means commonly used by those skilled in the art, and will not be described in detail herein.
  • Each CCD sensor has a maximum operating range, referred to herein as exposure time. Above a certain maximum input value, the output signal will no longer increase and the sensor will go into saturation. In addition, the sensor also has a minimum response value below which the sensor does not respond.
  • a white uniform target plate is placed on the sample stage in the black chamber. This is done to estimate the first appropriate value for the exposure time of each filter.
  • the average value of the captured image is set to obtain a predetermined value s.
  • This predetermined s is close to the maximum possible value or saturation value of the signal (for example, 16384 for 14-bit quantization), but in order to avoid saturation or supersaturation of the acquired image, a safe interval is usually set for the predetermined value s.
  • a multiplication correction factor such as 0.75 is typically applied to estimate the exposure time for each channel.
  • Figure 9 is a graph showing the relationship between the average response value of the captured image in each channel and the respective exposure time intervals. As shown in Figure 9, the CCD sensor operates in a linear dynamic range with various maximum estimated time intervals between 35ms and 340ms.
  • the CCD must perform four tasks in the image generation process, including: charge generation, charge collection, charge transfer, and charge measurement.
  • the charge generation for each pixel is proportional to the current incident light level, so the interaction of all pixels produces a spatial sample representation of a continuous image.
  • the CCD sensor accurately reproduces the image after generating the electrons.
  • the reproduced digital image includes an electronic charge pattern for each pixel, the electronic charge pattern being represented by an array holding electrons generated during integration.
  • no light reaches the detector of the CCD the charge of each pixel is incrementally transferred between pixels.
  • ADC analog-to-digital converter
  • the noise source inherent in an imaging system changes its digital level corresponding to each pixel, and also causes distortion of the real image and degradation of radiation accuracy, image quality, and resolution.
  • the most important source of noise is dark current.
  • the generation of dark current noise is a thermal process in which the electron endotherm steps to an intermediate state, that is, it is excited into the conduction band. For this reason, the most effective way to reduce dark current is to cool the CCD.
  • the CCD sensor operates below minus 30 degrees, and a cooling system can be designed in the light box to reduce the ambient temperature to maintain the steady state of the multispectral imaging color measurement system.
  • a stable and uniform light source is another important factor for obtaining a high quality image for photometric measurement.
  • a non-uniform incident source will result in small changes in the pixel response of the image system in the visible range, affecting the results of the color measurement.
  • a slightly different detector size and doping concentration will result in slightly different amounts of dark current per pixel, which is another major source of dark current non-uniform noise.
  • the multi-spectral imaging color measurement system of the present invention not all pixels in the CCD sensor have the same light sensitivity. Even small changes in the thickness of the silicon wafer will affect sensitivity.
  • the light incident on the sensor may not be uniform due to light loss occurring in the optical lens. These small changes due to changes in the brightness of the object in the image itself cannot be detected, and these unwanted pixel or illumination variations can affect the measurement accuracy of the multispectral imaging color system to some extent.
  • Ic. Rr stands for the corrected image
  • I rep stands for the original or original uncorrected image
  • I Dark stands for the "dark current” image
  • I wh stands for the uniform white target image
  • the coefficient k is a guarantee
  • the calibration constant of the CCD sensor operating in the linear range state is acquired by a black-and-white CCD camera with a filter wheel, and the different refractive index of the filter has a certain displacement offset due to the different refractive index of the filter.
  • an image of a single corrected object (e.g., a white-black checkerboard pattern) is acquired in the designed black room.
  • select an image of a channel as a reference image such as an image on a 560 channel.
  • the matching corrections for the images of all other channels are relative to the selected reference channel image.
  • a multi-channel image calibration algorithm is used to obtain a multi-spectral calibration image of the sample being tested.
  • the images of all other channels are calibrated based on the selected reference channel image.
  • an appropriate threshold is selected to binarize them.
  • the multi-spectral image is then screened by a gradient of local regions using an edge detection method. Since all localized regions of the multispectral image retain feature edges, the edge selection of the input image provides robustness over a wide range of wavelengths.
  • the capture step of the multi-channel image calibration method is shown: the measured pattern is a black and white grid pattern, the reference channel (560 nm wavelength) image is shown in Fig. 11B, and the target channel (700 nm wavelength) image is shown in Fig. 11B. .
  • Figure 11C shows the difference between the pre-calibration reference channel image and the target channel image
  • Figure 11D shows the difference between the calibrated reference channel image and the target channel image. It can be seen that this difference is significantly reduced after calibration.
  • the geometric distortion in the X and y directions is neither a spatial constant on a certain channel nor a spatial constant of an image on the same channel of a different object. In fact, it depends on the object distance, camera zoom and aperture, so the software is recalibrated each time you have a multi-spectral exposure.
  • the reference image and the target image are divided into a series of partitions to take into account the unevenness of the image distortion vector. Each interval is calculated separately when calculating the distortion vector.
  • mapping function f ⁇ ', which converts the spatial coordinate X of the target image ⁇ into the coordinate ⁇ of the reference image S.
  • the method of selecting the mapping function f is as follows: taking the minimum value of the error cost function of the spatial displacement, and then finding the maximum value of the correlation coefficient between the edge portion of the target image and the corresponding portion of the reference image.
  • the mathematical formula for the multi-channel image calibration process is as follows: max /(5( (x), 7( ), )) where I ( ) represents the selected cost function. Find the maximum correlation of each interval distortion vector with a suitable algorithm Coefficient.
  • the maximum displacement of the edge of the target image and the reconstructed image in both the horizontal and vertical directions is used to record all other channel images.
  • a vector group of the same size as the original image is generated.
  • the set of vectors includes a distortion vector for the selected interval. Bilinear interpolation of portions other than edges in the target image generates a vector of remaining pixels.
  • filters with light reflection and transmission interference in multispectral imaging color measurement systems are not ideal optical components.
  • the filter is coated with an anti-reflection film, a portion of the incident light is reflected on the surface of the medium. Further, as shown in Figure 12, the two media surfaces of the filter are not coplanar. In addition, there will be single or multiple reflections between the filter and the lens system. These non-ideal optical properties result in undesirable dual imaging or superimposing effects in the acquired image.
  • the aliasing effect of an imaging channel is usually different from that of other channels because this effect is mainly caused by non-ideal design and filter fabrication processes.
  • the ghosting effect will undoubtedly affect the response of the imaging system (such as the camera) at each pixel location, reducing the accuracy of the color measurement of the multispectral imaging color measurement system.
  • the ratio of the brightness of the overlay to the brightness of the object varies from channel to channel, typically less than 2°/. ,. This density ratio will severely reduce the accuracy of luminosity and chromaticity, especially for samples with low brightness, so the ghosting effect on each imaging channel is corrected.
  • Luminance ratio and positional shift are two important parameters that cause the ghosting effect.
  • a planar sample of a white cross shape in a dark background was imaged. It should be understood that the white object is not limited to the shape of the cross, and other shapes are also possible. Due to the superimposing effect, in addition to the real object, the resulting image also includes a ghosting cross frame with 4 inches of low brightness. The position of the ghost cross is determined by the template matching method.
  • the extraction of white objects is achieved by means of image thresholds.
  • Pixels with brightness greater than T are considered to be in the candidate object, while other pixels are in the background or in the overlay. Due to the influence of image noise, there may be isolated pixels or tiny intervals that are considered candidates. After the position of the largest candidate object is determined, a partial image containing the real object can be identified, that is, the template image I
  • Figure 14 shows the process of determining template matching for the position of the overlay.
  • the matching process runs in scan mode, for example, from left to right or top to bottom.
  • the following equation gives the correlation coefficient between the template starting from the position (s, t) and the candidate partial graph. It can be seen that if the pixel (s+m, t+n) is within the object area, the pixel will not be calculated because it is the object pixel.
  • the marks ⁇ and in the above equation represent the average brightness of the object portion and the candidate overlap portion, respectively.
  • the luminance ratio of the ghosting effect is calculated by:
  • r B , r c , I are the average brightness of the regions B, C and D, respectively.
  • the calibration density of the pixel position (i, j) is:
  • I ⁇ (i, j) I(i,j)- ⁇ -I(i+i offset , j + j offset )
  • ⁇ and (i. ffset , j. ffset ) refer to a single filter
  • the elimination of the ghosting effect is performed separately according to different imaging channels.
  • the primary goal of the multispectral reconstruction method is to reconstruct the reflectance spectra of the color samples from the corresponding digital responses of the imaging system. Reflectance reconstruction methods are commonly used in multispectral imaging systems because the linear model used requires the acquisition of more spectral channels to estimate reliable spectral reflectance.
  • the mathematical methods of spectral reconstruction can be divided into interpolation methods, such as Lagrangian polynomial interpolation, cubic spline interpolation, cubic interpolation, discrete Fourier transform or modified discrete sine transform, and estimation methods, such as pseudo-inverse method, smoothing Pseudo-inverse, Wiener estimation, nonlinear methods, principal component analysis, independent component analysis, or non-negative matrix factorization.
  • interpolation methods such as Lagrangian polynomial interpolation, cubic spline interpolation, cubic interpolation, discrete Fourier transform or modified discrete sine transform
  • estimation methods such as pseudo-inverse method, smoothing Pseudo-inverse, Wiener estimation, nonlinear methods, principal component analysis, independent component analysis, or non-negative matrix factorization.
  • the estimation method is usually based on the known knowledge of the type of spectrum found in a series of measurements previously implemented, ie the training set.
  • Wiener estimation or pseudo-reverse calibration is used for spectral reconstruction. This is in the context of background technology and can be found in the relevant publications, which is not described here.
  • the calibrated image data can be used to measure the color spectrum of each image point of the measured object with high photometric accuracy.

Landscapes

  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)
  • Spectrometry And Color Measurement (AREA)

Description

多光讲成象顧色测量系统及其成^ ί象信号处理方法 技术领域
本发明涉及成像领域,尤其涉及一种多光谱成像颜色测量系统及其成像信 号处理方法。 背景技术
目前, 用分光光度计可以对大型单色样品进行颜色测量, 并具有较高颜色 精确度。 但是, 由于分光光度计的空间分辨率极低, 它不能用于测量各种纺织 及成衣样品, 例如色织面料、 蕾丝、 饰品和印花面料。
传统成像设备包括数码相机和彩色扫描仪,它们记录颜色信息的方法是物 体表面反射光通过具有不同光谱透过率的多个颜色滤光片,然后利用电子传感 器将所述反射光形成图像。 它们在同时测量很多颜色方面具有很多优点,在测 量小型、 多色物体和捕捉具有复杂纹理表面的物体的颜色信息时尤其常用。 然 而, 一方面, 传统三原色和色度法成像系统由于违背了 Luther规则而很难产 生高颜色精确度的图像, Luther 规则是指多数相机的光 if^丈感度并不与人眼 相似或并是不是人眼的线性组合。 另一方面, 这种色度法成像系统仅能输出三 刺激值, 这种方式会导致由不同观察者和不同照明系统引起的同色异谱错误。 因此, 这些色度法测量系统, 例如由 Ver ivide有限公司生产的 Dig iEye, 对 于颜色质量要求较高的纺织及成衣工业, 其精确度不够。 发明内容
本发明要解决的技术问题在于,针对现有技术中的多光谱成像颜色测量系 统在应用于纺织及成衣工业中精确度不够的缺陷,提供一种高精确度的多光谱 成像颜色测量系统及其成像信号处理方法。
本发明解决其技术问题所采用的技术方案是:
提供一种多光谱成像颜色测量系统,包括用于为多光谱成像形成包围密闭 空间的黑室、位于黑室内用来安放和固定被测样品的载样台以及位于黑室内用 来拍摄被测样品图像的成像系统, 还包括可控照明系统、 滤光片轮系统、 成像 信号处理系统和电子控制系统,
所述可控照明系统位于所述黑室内,包括至少一个在被测样品上侧对称排 布并指向所述被测样品的灯管;
所述滤光片轮系统位于所述成像系统与所述被测样品之间,用于过滤经所 述被测样品反射的所述可控照明系统发出的光;
所述成像信号处理系统位于所述成像系统内,用于对所述成像系统拍摄的 图像进行校准和反射重构;
所述电子控制系统与所述可控照明系统、所述滤光片轮系统以及所述成像 系统通信相连, 用于控制上述各个系统的运行状态。
本发明一种多光谱成像颜色测量系统中, 所述灯管包括依次连接的光源、 用于收集更多均匀光的光学积分柱、用于进一步改善照明均匀性和放大倍数的 透镜组以及用于减少漫射光的光阻板,所述灯管的内部还涂覆有用于减少内部 漫射光的吸光材料;
所述光源是具有平滑曲线光谱能量分布的面钨灯,由两个高精度直流稳压 电源供电;
所述光学积分柱是由玻璃墙包围的中空喇叭结构;
所述透镜组包括用于限制光束边缘的光栅,还包括一个或多个具有不同折 射率的凸透镜和凹透镜;
所述光阻板置于所述透镜组的前边缘。
本发明一种多光谱成像颜色测量系统中,所述滤光片轮系统包括电机、滤 光片轮以及连接所述电机和所述滤光片轮的皮带装置;
所述电机用于给所述滤光片轮提供用于滤光片位置选择的电源和通讯,由 所述电子控制系统控制;
所述滤光片轮进一步包括底盘,所述底盘上设有一个或多个插槽以及一个 或多个用于减小旋转负重的洞, 滤光片通过顶扣环固定在所述插槽中, 所述滤 光片轮还包括一个内置的用于位置探测的红外光学开关;
所述皮带装置进一步包括装在所述电机的旋转轴上的第一同步轮、装在所 述底盘的轴上的第二同步轮以及与所述第一同步轮和第二同步轮外边缘齿合 的皮带。
本发明一种多光谱成像颜色测量系统中, 所述成像系统进一步包括 CCD 传感器或 CMOS传感器, 所述传感器内置有 A/D转化器, 用于将投射到所述传 感器焦平面上的经过所述滤光片轮系统选定波长的光信号转换为数字电信号, 产生多通道光谱图像。
本发明一种多光谱成像颜色测量系统中, 所述电子控制系统包括: 用于调整所述照明系统电源的电压和电流设置值以符合光源稳定工作状 态的照明系统控制单元;
用于控制所述电机的加速、稳态驱动和减速这三种工作状态的微处理器单 元;
用于所述成像系统、照明系统控制单元和滤光片轮系统间通讯的接口电路 板。
本发明还提供一种成像信号处理方法,包括对光学透镜或滤光片引起的误 差进行几何校正, 及对每个光通道图像进行亮度校正, 还包括: 估算传感器线 性工作范围的曝光时间,使所述传感器工作在所述曝光时间内以便将入射光信 号转化为数字成像信号;对所述传感器中固有噪声源引起的图像噪声进行校正; 对于不同滤光片下获取的不同光通道的图像进行多通道图像配准,以消除由于 滤光片轮中滤光片的相对位置差异、滤光片的不同折射率影响、镜头色散或成 像系统与被测物体间的微小差距造成的不同通道图像间的内容偏移;对每个通 道上的叠影效应引起的叠影图像分别进行校正;
利用经过上述处理后的成像信号重构反射率,生成所述被测样品的光谱反 射率图像。
本发明一种成像信号处理方法中,所述固有噪声源包括暗电流, 用 Ic。rr代 表校正后的图像, Irep代表最初或原始的没有校正过的图像, IDark代表 "暗电 流" 图像, Iwhits代表均匀白色目标图像, 系数 k是一个保证 CCD传感器工作在 线性范围状态的校准常数, 则校准过程用下式表示: 本发明一种成像信号处理方法中, " 对于不同滤光片下获取的不同光通道 的图像进行多通道图像配准" 进一步包括:
选择一个通道的图像作为参考通道图像,根据选定的参考通道图像分别对 其它目标通道图像进行校准;
对采集的各通道图像分别进行二值化预处理,然后利用边缘检测算法提取 各通道图像的特征边缘;
将所述参考通道图像和目标通道图像分成一系列局部区域,选择空间位移 的误差代价函数,取其最小值, 用基于梯度下降算法求所述每个局部区区域内 失真向量的最大相关系数;
将目标图像中除边缘部分外的其它部分进行双线性插值生成局部区域内 其他像素的偏移向量,将所述剩余像素的偏移向量与所对应的局部区域的边缘 部分的偏移向量一起生成所述选定局部区域内与原始图像相同大小的偏移向 量的向量组, 产生映射函数 f (x) ;
根据所述生成的映射函数 f (X)对所述目标通道图像进行恢复校准。
本发明一种成像信号处理方法中, "对每个通道上的叠影效应引起的叠影 图像分别进行校正" 进一步包括:
利用图像阈值方法提取白色物体;
以扫描方式对整个图像的每个像素进行模板匹配处理,确定叠影效应造成 的叠影图像的位置;
对于每个成像通道的叠影图像分别根据其相关参数消除叠影效应。
本发明一种成像信号处理方法中, "利用经过上述处理后的成像信号重构 反射率" 具体包括使用维纳估计或伪逆校准来重构反射率。
本发明多光谱成像颜色测量系统及其成像信号处理方法的有益效果是:多 光谱成像颜色测量系统可以克服传统数码成像系统和分光光度计系统的缺点。 在均匀照明系统设计、多光谱成像信号处理系统和多光谱成像信号处理方法的 帮助下, 所述完整的多光谱成像颜色测量( IC匪)系统可以为纺织工业应用用 户提供颜色测量和评定的基本功能。本发明设计了均匀照明系统来提供均匀的 成像环境,并设计了成像信号处理系统以便同步捕捉高空间分辨率的光谱颜色 图像。 因此, 任意多种颜色, 无论该颜色多小, 本发明都可以轻易测量。 这种 技术在提供物体的结构信息方面优于分光光度计,且在获取真实光谱反射和消 除不同光源下的同色异谱方面优于现在的数码相机系统。 附图说明
下面将结合附图及实施例对本发明作进一步说明, 附图中:
图 1是根据本发明一个实施例的成像颜色测量系统的示意图;
图 2是根据本发明一个实施例的灯管的示意图;
图 3是根据本发明一个实施例的光学积分柱的示意图;
图 4是根据本发明一个实施例的透镜系的示意图;
图 5是根据本发明的一个实施例的含两个灯管的对称照明系统的示意图; 图 6是根据本发明的另一个实施例的由八个成 45°均匀分布的光源组成的 环状照明系统的示意图;
图 7A是根据本发明的另一个实施例的多层结构照明系统的示意图; 图 7B是根据本发明的另一个实施例的多层结构照明系统的夹角的示意图; 图 8是根据本发明的一个实施例的滤光片轮系统的示意图;
图 9是根据本发明的成像信号处理方法的流程图;
图 1 0是每个通道中拍摄图像的平均响应值与各自曝光时间区间的关系曲 线图;
图 11A是多通道图像校准过程中的参考通道图像;
图 11B是多通道图像校准过程中的目标通道图像;
图 11C是多通道图像校准前的参考通道图像与目标通道图像的位置偏差 示意图;
图 11D是多通道图像校准后的参考通道图像与目标通道图像的位置偏差 示意图;
图 12是根据本发明一个实施例的成像颜色测量系统中滤光片的非平行表 面示意图;
图 1 3是根据本发明一个实施例的叠影效应示意图; 图 14是根据本发明一个实施例的模板匹配过程示意图。 具体实施方式
图 1是根据本发明一个实施例的成像颜色测量系统的示意图。如图 1所示, 为了避免外部环境光线对颜色测量的干扰, 所有的成像部件都装在黑室 6中, 将白板 5和样品放在两个对称的分别以 45 士 5° 的几何角度倾斜的灯管 4下。 为了达到所要求的均匀的光强分布,对可控照明系统进行了特殊设计。成像系 统 1、 分光片轮 3和透镜 2置于顶盒中, 用于实现物体的多光谱成像。 本发明 多光谱成像颜色测量系统中使用的是窄带干涉滤光片。
图 2是根据本发明的一个实施例的灯管的示意图。 在每个灯管中, 包括 3 个单元,例如光源、光学积分柱和透镜系。面钨灯用作光源,可以发出持续的、 光谱功率集中分布在可视光长范围内的、在一定时间内光强稳定的光。 光学积 分柱用于聚集尽可能多的光线,它可以在光学积分柱出口产生均匀扩散的反射 光,所述反射光再通过透镜组以进一步改善均匀照明有效范围内的均匀性和放 大倍数。 最后, 可以提供达到多光谱成像颜色测量系统要求的均匀照明区域。
高压氙灯也能发射持续光谱, 它的色温一般在 5000k。 如果将某种转换滤 光片与氙灯结合,可以更好的模拟 D65标准的照明体。但是就像所有其它气体 放电光源一样, 氙灯的发射光谱有特定的谱线, 它的波长大概在 475nm左右。 为了减小特征谱线的影响,通常使用特殊的分光片来抑制发射线谱。对比来说, 卤钨灯的相应光谱能量分布是没有尖峰或小抖动的平滑曲线。在本发明中,用 卤钨灯作为光源, 为多光谱成像颜色测量系统提供持续光谱分布的光。
作为多光谱成像颜色测量系统的光源,平面光源的稳定性是获取高质量图 像的第一步。 两个高精度直流稳压电源为面钨灯提供能稳定电压和电流, 以保 证光源光强的稳定性。有线或无线接口用于连接直流稳压电源和主机。主机发 出指令,根据规定的一些有线或无线协议,调整所需的电压和电流值以符合光 源稳定工作状态的设置值。
另外, 为了进一步减小光路和灯管中漫射光的影响,每个灯管中进一步包 括光阻板。光阻板置于每个透镜组的前边缘, 用来减少所照明物体表面的漫射 光。 为了进一步减少漫射光, 整个灯管内部还可以涂抹吸光材料。 图 3是根据本发明一个实施例的光学积分柱的示意图。如图 3所示,将卤 钨灯中经过多重反射的发射光看作一个具有一定规模和高亮度的点光源。这个 点光源的亮度分布类似于高斯分布, 它是不均匀的。 为了达到多光谱成像颜色 测量系统中要求的 20cm*20cm区域内的均匀照明条件,要对所述发射光作进一 步处理。 在本发明中, 设计了一个重要的光学器件来导光, 即光学积分柱。 它 的主要作用就是收集更多的均匀光。
在本发明中,积分柱是由光学玻璃片包围的中空结构,用来均匀化发射光。 这种喇叭结构用来收集尽可能多的光,例如可以获取光学积分柱左侧的平面漫 反射光。 当导入积分柱的光束的扩散角小于光导的孔径角时, 经玻璃内壁多次 反射的光将从积分柱的出口射出。 当光束的扩散角大于光导的孔径角时,所述 光束将发生折射。通过这种光导机制,从积分柱另一端出来的光将形成强度更 均匀的光束, 而这时光的传播方向是杂乱的。 光折射管孔径角由它的折射系数 决定。
图 4是根据本发明一个实施例的透镜系的示意图。尽管从光学积分柱中经 多次镜面反射出来的光可以看作是均匀的,但这种光的均匀照明区域的有效范 围相对较小,不能满足多光谱成像颜色测量系统中测量平面上大规模均匀照明 区域的要求。 因此, 在本发明中, 使用一系列透镜来扩大从光学积分柱中出来 的光的均匀照明区域。 为了保证透镜系的放大系数, 减少光学畸变、 透镜像差 和色差, 本发明采用的透镜系包括多片透镜, 具体光路图如图 4所示。
本发明的一个实施例所采用的透镜系包括两个透镜组。第一组包括五片透 镜, 其中有两片凹透镜和三片凸透镜。 第二组包括一片凹透镜和一片凸透镜。 将凹透镜和凸透镜组合使用是为了消除透镜组的几何畸变。透镜系中选择不同 折射率的透镜是为了消除色散误差。如图 4所示,透镜系中的光栅的主要作用 是限制光束边缘以免投射到外部。在一般情况下, 光束边缘的畸变甚至比改变 透镜位置造成的畸变还要严重。所述光栅可以有效地增加屏幕亮度均勾性和对 比度, 但同时也限制了光源的利用率。
图 5是根据本发明的一个实施例的含两个灯管的对称照明系统的示意图。 如图 5所示, 两个灯管对称的分别以 45 ± 5° 的几何角度倾斜放置。 尽管图 1 和图 5中给出的照明系统采用了两个对称的灯管,但并不限制于两个对称灯管, 在本发明的另一个实施例中,本发明的环状照明系统可以包括任意数量或组合 的卤钨灯光源。 另外, 本发明的照明系统还可以包括任意层数的光源, 以产 生均匀光照区域。
例如,图 6示出了根据本发明的另一个实施例的环状对称照明系统的示意 图。 如图 6所示, 八个光源相互间隔 45°均匀分布在圆环上。
图 7A和图 7B进一步示出了多层环状照明系统的示意图。图 7中每层的每 个光源都以圆圈形式筒单表示。这种多层环状照明系统的每一层都能独立工作。 多层环状照明系统可以从不同角度照射所测量的物体,以获得对定向纹理图案 样品和其它颜色依赖于光源不同入射角效果的物体的更精确的光谱分析。
图 8是根据本发明的一个实施例的滤光片轮系统的示意图。 如图 8所示, 滤光片轮包括许多安装在轮齿上的窄带滤光片,窄带滤光片在轮上连续紧密排 布。这些波长可选的窄带滤光片允许不同波长范围的光通过相应的滤光片。例 如,这些波长可选窄带滤光片的不同区域可以将光过滤为 10或 20纳米宽的不 同光谱。例如,所述波长可选窄带通滤光片的不同区域可以将光过滤为分别以 400nm、 420nm、 440nm、 460nm、 480nm、 500nm、 520nm、 540nm、 560nm、 580nm、 600謹、 620謹、 640謹、 660謹、 680謹 和 700謹为中心的不同的 20謹带宽的 光谱。 在本发明中, 窄带滤光片的数量不限于 16或 31 , 滤光片的带宽也不限 于 20謹或 10nm。
图 8示出了本发明的一个实施例, 该实施例中的滤光片轮包括 16个窄带 滤光片。在这个实施例中, 安装有波长可选窄带滤光片的自定义滤光片轮可以 用在成像系统和被测物体之间来采集每个光谱通道的信息,使成像系统用作彩 色分析仪 /传感器。 所述窄带滤光片为光谱分析和颜色测定提供合适波长范围 的光谱。 例如, 波长可选窄带滤光片可以提供 16个不同波长范围的光谱以供 探测和分析, 当然, 也可以是其它适当数目的滤光片或光谱。
本实施例所采用的滤光片轮包括一个底盘, 底盘上有 16个用于安置滤光 片的插槽。 每个插槽中, 用一个顶扣环来固定滤光片的边缘, 减小滤光片轮旋 转时滤光片的倾斜。在滤光片轮中, 用一个皮带装置来取代直接耦合到步进电 机驱动器轴上的方式, 以便携带较重的负载, 并保证更高的转动精度。 第一同 步轮装在步进马达的旋转轴上, 第二同步轮装在轮底盘轴的外围,在轮底盘的 中心轴和外围之间装有许多等距环状排布的窄带滤光片。轮底盘上各种各样的 洞用于减小其旋转时的负重。与第一同步轮和第二同步轮各自外边缘齿合的皮 带的作用是, 以第一同步轮为中心, 带动滤光片轮底盘旋转。
图 8所示的滤光片轮是一个一端接有步进电机的紧凑圆盘设计。所述滤光 片轮有一个内置的用于位置探测的红外光学开关,可以在每次旋转开始时对轮 子进行初始位置校准。 步进电机由微处理器模块控制,通过有线或无线接口端 将滤光片轮和主机连接起来。步进电机是一个给电动滤光片轮提供用于滤光片 位置选择的电源和通讯的外部元件。在本发明中,设计了两个互补控制来检测 滤光片轮机制的旋转速度和位置。使用能给出参考位置的增量编码器就能实现 自动控制。 步进电机即使停止时通常也是通电的, 固定在插槽中时没有明显的 反弹或位置漂移。
当轮子旋转时, 微处理器将三种工作状态的旋转脉沖发给电机的驱动器, 所述三种工作状态包括加速、稳态驱动和减速,同时,将启动频率、驱动频率、 加速时间和减速时间设置为相应的预定值。在滤光片轮的每个静止位置, 光束 以选定的波长通过滤光片和光学透镜传播。然后,投射到电荷耦合器件( CCD ) 或互补金属氧化物半导体( CMOS )传感器的焦平面上, CCD或 CMOS传感器可 以通过内置于成像传感器电路的 A/D转换器将模拟光信号转换为数字电信号。 步进电机带动轮子连续旋转, 成像传感器就会产生 16个或更多通道的光谱图 像。
为了便于成像系统与照明系统控制单元和滤光片轮间的通讯,安装了一个 接口电路板。该接口可以支持任意合适类型的通信媒介, 例如有线或无线的网 络或连接 。 该接口可以包括任意适合微处理器与主机间通信的结构。 例如, 可以是支持相机曝光时间的同步机制的接口电路板,也可以是支持窄带滤光片 按照 RS232通信协议连接的持续定位的接口电路板。
图 9是根据本发明的成像信号处理方法的流程图。如图 9所示,将原始图 像转换为光度成像信号的成像信号处理方法包括: 曝光时间估算、成像噪声校 正、、 多通道图像校准、 叠影校正、 图像亮度校正以及反射率重构。 移动镜头 的几何校正和亮度校正是本领域技术人员常用的技术手段,本文不再进行详细 描述。
每个 CCD传感器都有最大工作范围,在本文中称为曝光时间。超过某个最 大输入值, 输出信号将不再增加, 传感器进入饱和。 此外, 所述传感器还有一 个最小响应值, 低于这个值, 传感器不做出响应。 为了估算 CCD传感器在每个 通道的曝光时间的线性范围, 在黑室中的样品台上放置一个白色均匀目标板。 这样做是为了估算每个滤光片的曝光时间的第一个适当值。
作为曝光时间估算的基本标准,设置拍摄图像的平均值,来得到某个预定 值 s。 这个预定 s 接近于信号的最大可能值或饱和值(例如 14 位量化时是 16384 ),但为了避免出现采集图像出现饱和或过饱和的情形,通常会对预定值 s设定一个安全区间。 例如, 通常将乘法修正系数例如 0. 75应用于估算每个 通道的曝光时间。
图 9 是每个通道中拍摄图像的平均响应值与各自曝光时间区间关系的曲 线图。 如图 9所示, 采用介于 35ms到 340ms之间的各种不同的最大估算时间 区间可以保证 CCD传感器工作在线性动态范围。
从技术上来说, CCD必须执行图像生成过程中的四个工作, 包括: 电荷产 生、 电荷收集、 电荷转移和电荷测量。 每个像素的电荷生成与当前入射光水平 成比例, 因此所有像素的共同作用就生成了一个连续图像的空间样品代表。在 电荷收集过程中, CCD传感器在生成电子后对图像进行精确的再现。 再现的数 码图像包括每个像素的电子电荷模式,所述电子电荷模式是以保存有整合期间 产生的电子的数组来表示的。 当没有光到达 CCD的探测器时,每个像素的电荷 在像素间递增转移。 最后, 在电荷测量阶段, 每个像素的电荷量都连接到输出 放大器, 然后通过模数转换器(ADC )按顺序数字化。
在生成图像的过程中, 成像系统(例如 CCD相机)中固有的噪声源会改变 其相应于每个像素的数码水平,还造成真实图像的畸变以及辐射精度、 图像质 量和分辨率的下降。 在本发明中, 最重要的噪声源是暗电流。
暗电流噪声的产生是一个热过程, 其中, 电子吸热阶跃到一个中间状态, 即被激发进入导带。 由于这个原因, 减小暗电流的最有效的方法就是给 CCD 降温。 在本发明中, CCD传感器工作在零下 30度以下, 在灯盒中还可以设计 了一个降温系统来降低环境温度,以维持多光谱成像颜色测量系统的稳定状态。
尽管通过在硬件部分设计中减小光强的波动可以很大程度上改善光源的 重复率,一个稳定和均匀的光源是获取用于光度测量的高质量图像的另一个重 要因素。 不均匀的入射光源将会导致可视范围内图像系统像素响应的小变化, 从而影响颜色测量的结果。
探测器大小和掺杂浓度稍有不同就会导致每个像素产生的暗电流数量稍 有不同, 这是另一种主要的暗电流不均匀噪声信号产生的原因。在本发明多光 谱成像颜色测量系统中,并不是 CCD传感器中所有的像素都有相同的光敏感度。 即使是硅晶片厚度的小变化也将影响敏感度。 另夕卜,尽管提供了一个可控照明 环境, 但由于发生在光学透镜中的光损失, 照到传感器上的光可能并不均匀。 由于图像本身里物体亮度的变化造成的这些小变化不能被检测到,这些不需要 的像素或照明变化会在一定程度上影响多光谱成像颜色系统的测量精度。上述 两种类型的空间不均匀共同造成了成像系统(例如 CCD相机)响应的空间不均 匀,如果基于 CCD相机的成像系统要实现高精度的幅度或光度测量, 就必须对 所述空间不均匀进行校正。
为了校正所述暗电流,使用黑暗背景或关闭快门或者两种方式一起作用来 拍摄黑暗图像,以消除由于黑暗图像匹配白色目标图像引起的曝光时间造成的 暗电流不均匀噪声。在这种情况下, 图像照明和设备响应的两种空间不均匀都 造成了图像不均匀。 校准的基本过程用以下等式表示:
j ― rep ― hark (j
1 white (i>j)― hark (}>])
其中, Ic。rr代表校正后的图像, Irep代表最初或原始的没有校正过的图像, IDark代表 "暗电流" 图像, Iwh 代表均匀白色目标图像, 系数 k是一个保证
CCD传感器工作在线性范围状态的校准常数。 用前置有滤光片轮的黑白 CCD相机获取多光谱图像, 由于滤光片的不同 折射系数导致所获取的不同光通道图像存在一定的位移偏移。
为了计算参考通道和其它通道图像间的空间位移,在所设计的黑室中采集 一个筒单的校正物体(例如白-黑棋盘格型) 的图像。 在这些多通道图像中, 选择某个通道的图像作为参考图像,例如在 560謹通道的图像。所有其它通道 的图像的匹配校正都是相对选定的参考通道图像的。
多通道图像校准算法用于获得被测样品的多光谱校准图像。所有其它通道 的图像是根据选定的参考通道图像校准的。为了准确计算各通道局部区域的空 间位移, 首先根据图像的灰阶直方图分布,选定一个合适的阈值将它们二值化 处理。 然后再利用边缘检测方法对多光谱图像进行以局部区域的梯度筛选。 由 于多光谱图像的所有局部区域都保留了特征边缘,对输入图像的边缘 选提供 了在宽波长范围的稳健性。
在图 11中, 示出了多通道图像校准方法的捕捉步骤: 被测图样是黑白格 子图样,图 11A中是参考通道( 560纳米波长)图像,图 11B中是目标通道(700 纳米波长)图像。 图 11C中是校准前参考通道图像和目标通道图像的差别, 图 11D中是校准后参考通道图像和目标通道图像的差别。 可以看出, 校准后这种 差别明显减小。
如图 11C所示,在 X方向和 y方向的几何失真既不是某一个通道上的空间 常量, 也不是不同物体同一通道上图像的空间常量。 实际上, 它取决于物体距 离、 相机变焦和光圈, 因此, 每次多光谱曝光时, 软件要重新进行校准。 为了 达到这个目的,将参考图像和目标图像分成一系列分区间, 以便考虑图像失真 向量的不均勾性。 计算失真向量时单独计算各个区间的。
多光谱图像校准的目的是产生映射函数 f : χ→χ ' , 将目标图像 Τ的空间 坐标 X转换为参考图像 S的坐标 χ, 。 在本发明中, 选择映射函数 f 的方法如 下: 取空间位移的误差代价函数的最小值, 然后求目标图像的边缘部分和参考 图像相应部分间的相关系数的最大值。 多通道图像校准过程的数学公式如下: max /(5( (x), 7( ), )) 其中 I ( )代表所选代价函数。 用合适算法找到每个区间失真向量的最大相关 系数。 除了参考图像的 560謹光谱外, 目标图像和重构图像在水平和竖直方向 的边缘最大位移都被用来记录所有其它的通道图像。 最后, 生成与原始图像相 同大小的向量组。所述向量组包括选定区间的失真向量。 目标图像中除了边缘 以外的其它部分的双线性插值生成了剩余像素的向量。 多通道图像校准后,在 本发明多光谱成像颜色测量系统中,偏差不超过一个像素, 能实现优良的空间 分辨功能。
由于设计和制作技术的限制,多光谱成像颜色测量系统中的具有光反射和 透射干扰的滤光片并不是理想的光学元件。 尽管给所述滤光片涂了防反射膜, 一部分入射光还是会在介质表面发生反射。 更进一步的, 如图 12所示, 滤光 片的两个介质表面不是共平面的。此外,在滤光片和透镜系之间会有单次或多 次反射。这些非理想光学特性导致在获得的图像中出现不希望的双重成像或叠 影效应。
某个成像通道的叠影效应通常和其它通道是不同的,因为这种效应主要是 非理想设计和滤光片的制作工艺引起的。 叠影效应毫无疑问将影响成像系统 (例如相机 )在每个像素位置的响应, 降低多光谱成像颜色测量系统的颜色测 量的精确度。 不同通道的叠影亮度与物体亮度的比率也不同, 一般低于 2°/。,。 这种密度比率将严重降低光度和色度的精确度, 尤其对于具有低亮度的样品, 因此每个成像通道上的叠影效应都要校正。
亮度比率和位置偏移是造成叠影效应的两个重要参数。为了确定这两个参 数, 如图 13所示, 对一个黑暗背景中的白色十字架形状的平面样品进行成像 处理。 应当理解, 白色物体并不限制于十字架形状, 其他形状也可以。 由于叠 影效应, 除了真实物体外, 所得到的图像还包括一个具有 4艮低亮度的重影十字 架。 重影十字架的位置由模板匹配方法确定。
计算叠影参数和消除叠影效应的方法描述如下:
1.白色物体的提取
白色物体的提取是通过图像阈值的方法实现的。阈值 T由图像的最大亮度 Imax和最小亮度 Imin决定: T =
2
还可以使用其它阈值方法。认为亮度大于 T的像素在候选物体中, 而其它像素 在背景或叠影中。 由于图像噪声的影响, 可能有孤立像素或微小区间被认为是 候选物体。确定了最大候选物体的位置后,就能识别包含真实物体的部分图像, 即模板图像 I
2.确定叠影效应的位置
图 14示出了确定叠影位置的模板匹配的过程。匹配过程以扫描方式运行, 例如, 从左到右或从上到下。 下式给出了从位置(s, t)开始的模板和候选分 图的相关系数。 可以看出, 如果像素(s+m, t+n)在物体区域内, 该像素将不 被计算, 因为它是物体像素。
∑∑(l(i0+m>j0+n)- rtemp)(l(s + m,t + n)— Tcmid )
77(S,t):
∑∑(I(i。 + j。 + n) - i mp)2 ∑∑(Ks + m,t + n)- T^)
上式中的标记 ^ 和 分别表示物体部分和候选叠影部分的平均亮度。
对整个图像的每个像素进行模板匹配处理,就可以找到具有最大相关系数 的位置(S。, t。)。 然后根据下式计算叠影的位移:
^offset ' Joffset) — — S0, Jo— 。 )
相应地, 叠影效应的亮度比率由下式计算:
Figure imgf000016_0001
其中, rB, rc, I分别是区域 B、 C和 D的平均亮度。
3.叠影效应的消除
得到叠影效应的参数后,就可以消除任意捕捉图像的叠影效应了。像素位 置(i, j ) 的校准密度是:
I~(i, j) = I(i,j)-^-I(i+ioffset, j + joffset) 注意参数 β和(i。ffset, j。ffset)指的是单个滤光片, 要根据不同成像通道分 别实行叠影效应的消除。 经过以上一系列成像处理,就可以高精度地利用数码图像值来估算反射重 构。多光谱重构方法的主要目标是重构来自成像系统的相应数字响应的彩色样 品的反射率光谱图。反射率重构方法通常用于多光谱成像系统, 因为所用的线 性模型需要采集较多的光谱通道才能估算可靠的光谱反射率。
光谱重构的数学方法可以分为插值方法,例如拉格朗日多项式插值、三次 样条插值、 三次插值、 离散傅里叶变换或修正离散正弦变换, 还有估算方法, 例如伪逆法、 平滑伪逆、 维纳估计、 非线性方法、 主成分分析、 独立成分分析 或非负矩阵分解。估算方法通常基于以前实现的一系列测量发现的光谱类型的 已知知识, 即训练集。
在本发明中,使用维纳估计或伪逆校准来进行光谱重构。这属于背景技术 范畴, 可以在相关出版物中找到详细描述, 在此不 #文介绍。
最后, 经过反射率重构,校准后的图像数据就可以用于测量被测物体每个 图像点的颜色光谱, 且具有高光度精确度。
虽然本发明是通过具体实施例进行说明的, 本领域技术人员应当明白,在 不脱离本发明范围的情况下,还可以对本发明进行各种变换及等同替代。另夕卜, 针对特定情形或材料, 可以对本发明做各种修改, 而不脱离本发明的范围。 因 此, 本发明不局限于所公开的具体实施例, 而应当包括落入本发明权利要求范 围内的全部实施方式。

Claims

权利要求
1、 一种多光谱成像颜色测量系统, 包括用于为多光谱成像形成包围密闭 空间的黑室、位于黑室内用来安放和固定被测样品的载样台以及位于黑室内用 来拍摄被测样品图像的成像系统, 其特征在于, 还包括可控照明系统、 滤光片 轮系统、 成像信号处理系统和电子控制系统,
所述可控照明系统位于所述黑室内,包括至少一个在被测样品上侧对称排 布并指向所述被测样品的灯管;
所述滤光片轮系统位于所述成像系统与所述被测样品之间,用于过滤经所 述被测样品反射的所述可控照明系统发出的光;
所述成像信号处理系统位于所述成像系统内,用于对所述成像系统拍摄的 图像进行校准和反射重构;
所述电子控制系统与所述可控照明系统、所述滤光片轮系统以及所述成像 系统通信相连, 用于控制上述各个系统的运行状态。
2、 根据权利要求 1所述的多光谱成像颜色测量系统, 其特征在于, 所述 灯管包括依次连接的光源、用于收集更多均匀光的光学积分柱、用于进一步改 善照明均匀性和放大倍数的透镜组以及用于减少漫射光的光阻板,所述灯管的 内部还涂覆有用于减少内部漫射光的吸光材料;
所述光源是具有平滑曲线光谱能量分布的面钨灯,由两个高精度直流稳压 电源供电;
所述光学积分柱是由玻璃墙包围的中空喇叭结构;
所述透镜组包括用于限制光束边缘的光栅,还包括一个或多个具有不同折 射率的凸透镜和凹透镜;
所述光阻板置于所述透镜组的前边缘。
3、 根据权利要求 1所述的多光谱成像颜色测量系统, 其特征在于, 所述 滤光片轮系统包括电机、滤光片轮以及连接所述电机和所述滤光片轮的皮带装 置;
所述电机用于给所述滤光片轮提供用于滤光片位置选择的电源和通讯,由 所述电子控制系统控制; 所述滤光片轮进一步包括底盘,所述底盘上设有一个或多个插槽以及一个 或多个用于减小旋转负重的洞, 滤光片通过顶扣环固定在所述插槽中, 所述滤 光片轮还包括一个内置的用于位置探测的红外光学开关;
所述皮带装置进一步包括装在所述电机的旋转轴上的第一同步轮、装在所 述底盘的轴上的第二同步轮以及与所述第一同步轮和第二同步轮外边缘齿合 的皮带。
4、 根据权利要求 1所述的多光谱成像颜色测量系统, 其特征在于, 所述 成像系统进一步包括 CCD传感器或 CMOS传感器, 所述传感器内置有 A/D转化 器,用于将投射到所述传感器焦平面上的经过所述滤光片轮系统选定波长的光 信号转换为数字电信号, 产生多通道光谱图像。
5、 根据权利要求 1所述的多光谱成像颜色测量系统, 其特征在于, 所述 电子控制系统包括:
用于调整所述照明系统电源的电压和电流设置值以符合光源稳定工作状 态的照明系统控制单元;
用于控制所述电机的加速、稳态驱动和减速这三种工作状态的微处理器单 元;
用于所述成像系统、照明系统控制单元和滤光片轮系统间通讯的接口电路 板。
6、 一种成像信号处理方法, 包括对光学透镜或滤光片引起的误差进行几 何校正, 及对每个光通道图像进行亮度校正, 其特征在于, 还包括: 估算传感 器线性工作范围的曝光时间,使所述传感器工作在所述曝光时间内以便将入射 光信号转化为数字成像信号;对所述传感器中固有噪声源引起的图像噪声进行 校正; 对于不同滤光片下获取的不同光通道的图像进行多通道图像配准, 以消 除由于滤光片轮中滤光片的相对位置差异、滤光片的不同折射率影响、镜头色 散或成像系统与被测物体间的微 d、差距造成的不同通道图像间的内容偏移;对 每个通道上的叠影效应引起的叠影图像分别进行校正;
利用经过上述处理后的成像信号重构反射率,生成所述被测样品的光谱反 射率图像。
7、根据权利要求 6所述的成像信号处理方法,其特征在于,所述固有噪声 源包括暗电流, 用 1<。„代表校正后的图像, Irep 表最初或原始的没有校正过 的图像, IDark代表 "暗电流" 图像, Iwh^代表均匀白色目标图像, 系数 k是一 个保证 CCD传感器工作在线性范围状态的校准常数, 则校准过程用下式表示: j ― ,reP
Figure imgf000020_0001
(
^ w hite ( — ark )
8、 根据权利要求 6所述的成像信号处理方法, 其特征在于, " 对于不同 滤光片下获取的不同光通道的图像进行多通道图像配准" 进一步包括:
选择一个通道的图像作为参考通道图像,根据选定的参考通道图像分别对 其它目标通道图像进行校准;
对采集的各通道图像分别进行二值化预处理,然后利用边缘检测算法提取 各通道图像的特征边缘;
将所述参考通道图像和目标通道图像分成一系列局部区域,选择空间位移 的误差代价函数,取其最小值, 用基于梯度下降算法求所述每个局部区区域内 失真向量的最大相关系数;
将目标图像中除边缘部分外的其它部分进行双线性插值生成局部区域内 其他像素的偏移向量,将所述剩余像素的偏移向量与所对应的局部区域的边缘 部分的偏移向量一起生成所述选定局部区域内与原始图像相同大小的偏移向 量的向量组, 产生映射函数 f (x) ;
根据所述生成的映射函数 f (χ)对所述目标通道图像进行恢复校准。
9、 根据权利要求 6所述的成像信号处理方法, 其特征在于, "对每个通 道上的叠影效应引起的叠影图像分别进行校正" 进一步包括:
利用图像阈值方法提取白色物体;
以扫描方式对整个图像的每个像素进行模板匹配处理,确定叠影效应造成 的叠影图像的位置;
对于每个成像通道的叠影图像分别根据其相关参数消除叠影效应。
10、 根据权利要求 6所述的成像信号处理方法, 其特征在于, "利用经过 上述处理后的成像信号重构反射率"具体包括使用维纳估计或伪逆校准来重构 反射率。
PCT/CN2011/078862 2010-11-01 2011-08-24 多光谱成像颜色测量系统及其成像信号处理方法 Ceased WO2012058977A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US13/882,508 US9417132B2 (en) 2010-11-11 2011-08-24 Multispectral imaging color measurement system and method for processing imaging signals thereof
EP11837503.9A EP2637004B1 (en) 2010-11-11 2011-08-24 Multispectral imaging color measurement system and method for processing imaging signals thereof

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201010539818.2 2010-11-01
CN201010539818.2A CN102466520B (zh) 2010-11-11 2010-11-11 多光谱成像颜色测量系统及其成像信号处理方法

Publications (1)

Publication Number Publication Date
WO2012058977A1 true WO2012058977A1 (zh) 2012-05-10

Family

ID=46023995

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/078862 Ceased WO2012058977A1 (zh) 2010-11-01 2011-08-24 多光谱成像颜色测量系统及其成像信号处理方法

Country Status (4)

Country Link
US (1) US9417132B2 (zh)
EP (1) EP2637004B1 (zh)
CN (1) CN102466520B (zh)
WO (1) WO2012058977A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103308466A (zh) * 2013-06-04 2013-09-18 沈阳仪表科学研究院有限公司 便携式滤光片色轮型多光谱成像系统及其光谱图像处理方法
CN115908174A (zh) * 2022-11-10 2023-04-04 深圳市中达瑞和科技有限公司 相对照度矫正方法、相关系统、设备和存储介质
CN119941815A (zh) * 2025-01-09 2025-05-06 武汉纺织大学 一种基于纹理特征加权校正的纱线照相测色方法及系统

Families Citing this family (73)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103940587B (zh) * 2013-01-21 2017-04-19 中国计量大学 一种光源光度和色度参数角分布特性的快速测量系统
CN103108107A (zh) * 2013-03-05 2013-05-15 广州广电运通金融电子股份有限公司 图像处理装置及其图像数据校正方法
CN103207015A (zh) * 2013-04-16 2013-07-17 华东师范大学 一种光谱重构方法及其光谱仪装置
CN103868851B (zh) * 2014-02-24 2016-03-30 北京空间机电研究所 一种基于陷波滤光片的多光谱相机成像系统
CN104019902B (zh) * 2014-06-16 2016-09-28 武汉工程大学 家用试纸阅读器装置及其检测方法
WO2016071909A1 (en) * 2014-11-06 2016-05-12 Ramot At Tel-Aviv University Ltd. Spectral imaging method and system
EP3035035B1 (en) 2014-12-18 2020-07-22 CLUTEX - Klastr Technické Textilie, o.s. A method of continuous measurement of colouring of textile surfaces and a measuring machine for carrying out the method
CN104655573A (zh) * 2015-01-22 2015-05-27 浙江大学 一种植物侧表面冠层的高光谱扫描系统
KR101718517B1 (ko) * 2015-04-03 2017-04-05 재단법인대구경북과학기술원 피부 헬스케어를 위한 모바일 다중 분광 이미징 디바이스, 사용자 모바일 기기 연결, 이미지 분석 방법
FR3036187B1 (fr) * 2015-05-12 2019-09-13 Commissariat A L'energie Atomique Et Aux Energies Alternatives Procede de correction d’une image de fluorescence
AU2016278201B2 (en) * 2015-06-15 2021-08-12 Agrowing Ltd Multispectral imaging apparatus
CN105107760B (zh) * 2015-08-24 2018-02-16 爱丁堡(南京)光电设备有限公司 一种主动激发式光致多光谱成像的分选设备
CN105306774A (zh) * 2015-11-05 2016-02-03 苏州威盛视信息科技有限公司 一种成像装置
TWI703313B (zh) 2015-12-09 2020-09-01 台灣超微光學股份有限公司 光譜儀的量測方法、光譜儀及其電子電路
CN105911534B (zh) * 2016-04-12 2018-11-20 中国科学院上海光学精密机械研究所 透过率随接收视场角变化的光学滤波装置
FR3052234B1 (fr) * 2016-06-07 2022-04-08 Tpl Vision Uk Ltd Dispositif et procede d'eclairage pour vision industrielle
KR101739696B1 (ko) * 2016-07-13 2017-05-25 서장일 재질인식 조명 시스템 및 이를 이용한 재질인식 방법
WO2018066212A1 (ja) * 2016-10-05 2018-04-12 シャープ株式会社 推定装置、および推定装置の制御方法
EP3528661B1 (en) * 2016-10-18 2022-03-23 Christina Wang Color reproduction services
ES2668941B1 (es) * 2016-11-23 2019-03-04 Medicion Control Y Regulacion S L Sistema para determinar el grado de destenido de telas textiles
US10444161B2 (en) * 2017-04-05 2019-10-15 Kla-Tencor Corporation Systems and methods for metrology with layer-specific illumination spectra
CN106932099B (zh) * 2017-05-16 2019-04-12 华中科技大学 一种多光谱色差在线检测仪及其应用
EP3639237B1 (en) * 2017-06-13 2024-02-21 X-Rite, Incorporated Hyperspectral imaging spectrophotometer and system
US10768339B1 (en) * 2017-07-25 2020-09-08 Southern Methodist University Imaging hidden objects
TWI635260B (zh) * 2017-09-05 2018-09-11 宏碁股份有限公司 校準顏色感應筆、電子系統及其操作方法
US10337923B2 (en) 2017-09-13 2019-07-02 Qualcomm Incorporated Directional interpolation and cross-band filtering for hyperspectral imaging
CN109916823A (zh) 2017-12-13 2019-06-21 香港纺织及成衣研发中心有限公司 多光谱颜色成像装置及校正方法
US11002676B2 (en) 2018-04-09 2021-05-11 Hunter Associates Laboratory, Inc. UV-VIS spectroscopy instrument and methods for color appearance and difference measurement
CN108844901B (zh) * 2018-06-25 2021-08-13 南京工程学院 多光谱图像采集装置
CN109164014B (zh) * 2018-06-28 2020-11-06 浙江理工大学 一种基于高光谱图像处理的多色织物润湿区域识别方法
US11336835B2 (en) 2018-10-04 2022-05-17 Barco N.V. Method and system for estimating exposure time of a multispectral light source
CN109431458A (zh) * 2018-12-21 2019-03-08 合肥奥比斯科技有限公司 多光谱光源及眼底成像系统
CN109459441B (zh) * 2019-01-15 2024-08-27 南京中禧智能科技有限公司 一种检测装置、系统及方法
CN109819150B (zh) * 2019-02-20 2022-03-15 深圳劲嘉集团股份有限公司 一种多通道图像采集装置以及采集多通道图像的方法
CN110018122B (zh) * 2019-04-26 2024-10-25 南京砺剑光电技术研究院有限公司 基于多光谱成像技术的迷彩伪装材料检测装置和方法
CN110132419B (zh) * 2019-05-20 2021-05-07 天津大学 文物色彩的色坐标测量方法及系统
CN110187093B (zh) * 2019-06-26 2025-02-18 上海艾瑞德生物科技有限公司 一种检测荧光免疫和胶体金试剂条的跨平台检测装置
JP7326972B2 (ja) * 2019-07-30 2023-08-16 株式会社リコー 表面特性評価方法、表面特性評価装置、及び表面特性評価プログラム
CN110501339B (zh) * 2019-08-13 2022-03-29 江苏大学 一种复杂环境下的布面定位方法
CN112396670B (zh) * 2019-08-15 2023-01-10 天津大学青岛海洋技术研究院 面向新型二进制图像传感器的图像重建方法
CN110660112B (zh) * 2019-09-29 2021-09-24 浙江大学 一种基于专用色卡和多光谱成像的绘画光谱重构方法
CN110658179B (zh) * 2019-10-09 2021-11-19 上海理工大学 基于多信号叠加和伪逆法的激光拉曼气体浓度检测方法
EP4127684A1 (en) * 2020-03-30 2023-02-08 Smartex Europe, Unipessoal Lda. Systems and methods for calibration
CN111314684B (zh) * 2020-04-13 2021-06-29 杭州国芯科技股份有限公司 一种基于同色异谱的白平衡校正的方法
CN111551266B (zh) * 2020-05-25 2021-01-22 吉林求是光谱数据科技有限公司 一种基于多光谱图像探测技术的环境色温测试方法及系统
FR3110994B1 (fr) * 2020-05-28 2022-08-05 Color Grail Res Procédé de reconstruction d’une image, notamment d’une image exacte en couleur, programme d’ordinateur, dispositif et système associés
CN111783578B (zh) * 2020-06-19 2024-11-26 无锡市金义博仪器科技有限公司 一种基于图像处理的高速光谱仪安全监测装置及监测方法
CN111879791B (zh) * 2020-07-30 2023-06-20 西湖大学 一种图案表面凸起特征增强的机器视觉系统及方法
CN111896489B (zh) * 2020-08-05 2023-06-30 曲阜师范大学 一种基于交叉偏振的六通道多光谱测量系统及方法
CN112161708B (zh) * 2020-09-30 2023-03-24 湖南福瑞印刷有限公司 用于颜色三刺激值计算的光谱预处理方法
CN114449137B (zh) * 2020-11-02 2025-06-13 北京小米移动软件有限公司 滤光片结构、拍摄方法、装置、终端及存储介质
CN112461833B (zh) * 2020-11-23 2024-04-19 广东弓叶科技有限公司 基于间歇滤光片轮机构的多光谱图像采集方法及系统
CN112730336A (zh) * 2020-12-25 2021-04-30 温州晶彩光电有限公司 一种基于颜色传感器的光谱反射率重建方法及系统
CN113125012B (zh) * 2021-03-26 2024-07-02 北京和众视野科技有限公司 一种深色纤维脱色显微系统和方法
CN113295387B (zh) * 2021-05-25 2023-11-17 中国科学院合肥物质科学研究院 一种多波段条状滤光片光学参数测试系统及其测试方法
CN113405662A (zh) * 2021-06-16 2021-09-17 顺德职业技术学院 一种光敏电阻的颜色识别装置及识别方法
CN113588222B (zh) * 2021-09-26 2022-01-11 武汉精创电子技术有限公司 一种墨色一致性检测装置及方法
KR102692570B1 (ko) * 2021-10-22 2024-08-06 삼성전자주식회사 이미지 센서의 스펙트럼 데이터를 처리하는 장치 및 방법
JP7760904B2 (ja) * 2021-12-14 2025-10-28 セイコーエプソン株式会社 分光撮影方法、分光撮影装置、及びコンピュータープログラム
DE102022102547A1 (de) * 2022-02-03 2023-08-03 Carl Zeiss GOM Metrology GmbH Verfahren und Messsystem zum dreidimensionalen Vermessen von Objekten
CN114813434B (zh) * 2022-02-16 2024-08-20 浙江工业大学 基于多目标优化和多光谱技术的水分和粒径同时检测的方法
CN114608705B (zh) * 2022-05-09 2022-07-29 湖北方圆环保科技有限公司 一种光谱信号数据采样和峰值检测方法
US12174071B2 (en) 2022-05-13 2024-12-24 Verity Instruments, Inc. System, apparatus, and method for improved background correction and calibration of optical devices
CN115079505B (zh) * 2022-05-31 2024-05-10 合肥埃科光电科技股份有限公司 基于泰伯效应的棱镜分光多光谱相机匹配校准装置及方法
US12165256B2 (en) * 2022-10-14 2024-12-10 Motional Ad Llc Color map layer for annotation
TWI876231B (zh) * 2022-11-22 2025-03-11 財團法人紡織產業綜合研究所 布料上色系統及布料上色方法
CN116109809A (zh) * 2023-02-01 2023-05-12 福建农林大学 一种显微多光谱成像的食用菌菌种表型参数采集装置及方法
CN115988334B (zh) * 2023-03-17 2023-06-13 江西北纬空间信息技术有限公司 一种自校正的数字化相机移动遥感系统及方法
IT202300015738A1 (it) * 2023-07-26 2025-01-26 Loro Piana S P A Sistema e metodo per controllare l’uniformità cromatica di una pluralità di porzioni di un articolo
CN116907646B (zh) * 2023-07-27 2025-11-21 天津天石户外用品股份有限公司 户外针织品染料色号分类方法及设备
CN119028476B (zh) * 2024-10-29 2025-02-21 杭州泽天春来科技股份有限公司 傅里叶红外气体分析仪的数据处理方法、系统及可读介质
CN120404644B (zh) * 2025-06-30 2025-09-02 苏州工学院 一种纺织纤维成分定性在线分析系统及方法
CN120823217B (zh) * 2025-09-18 2025-12-02 南昌师范学院 一种基于图像识别的橡胶地板硫化阶段异物检测系统

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1363826A (zh) * 2000-12-08 2002-08-14 格莱特格-曼克贝斯公开股份有限公司 平面被测对象的逐个像素光电测量装置
JP2005265752A (ja) * 2004-03-22 2005-09-29 Fuji Xerox Co Ltd 色彩計測装置および画像形成装置
CN1746659A (zh) * 2004-09-07 2006-03-15 三星电子株式会社 光学检测装置以及使用该装置的多通道试样分析器
US7079251B2 (en) * 2003-10-16 2006-07-18 4D Technology Corporation Calibration and error correction in multi-channel imaging
CN101193185A (zh) * 2006-11-30 2008-06-04 富士施乐株式会社 图像处理装置、图像读取装置及图像形成装置
US20100092083A1 (en) * 2008-10-10 2010-04-15 Xerox Corporation In-line linear variable filter based spectrophotometer
CN101863161A (zh) * 2009-03-25 2010-10-20 海德堡印刷机械股份公司 与角度相关的颜色值修正

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3762817A (en) * 1972-02-23 1973-10-02 Minnesota Mining & Mfg Spectral densitometer
US5850472A (en) * 1995-09-22 1998-12-15 Color And Appearance Technology, Inc. Colorimetric imaging system for measuring color and appearance
US6449004B1 (en) * 1996-04-23 2002-09-10 Minolta Co., Ltd. Electronic camera with oblique view correction
EP1213568B1 (de) * 2000-12-08 2005-12-28 Gretag-Macbeth AG Vorrichtung zur bildelementweisen Ausmessung eines flächigen Messobjekts
US6970240B2 (en) * 2003-03-10 2005-11-29 Applera Corporation Combination reader
US20050219659A1 (en) * 2004-03-31 2005-10-06 Shuxue Quan Reproduction of alternative forms of light from an object using digital imaging system
US20070161876A1 (en) * 2005-11-18 2007-07-12 Spectrx, Inc. Method and apparatus for rapid detection and diagnosis of tissue abnormalities
JP5075648B2 (ja) * 2008-01-15 2012-11-21 オリンパス株式会社 画像処理装置、画像処理プログラムおよび画像処理方法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1363826A (zh) * 2000-12-08 2002-08-14 格莱特格-曼克贝斯公开股份有限公司 平面被测对象的逐个像素光电测量装置
US7079251B2 (en) * 2003-10-16 2006-07-18 4D Technology Corporation Calibration and error correction in multi-channel imaging
JP2005265752A (ja) * 2004-03-22 2005-09-29 Fuji Xerox Co Ltd 色彩計測装置および画像形成装置
CN1746659A (zh) * 2004-09-07 2006-03-15 三星电子株式会社 光学检测装置以及使用该装置的多通道试样分析器
CN101193185A (zh) * 2006-11-30 2008-06-04 富士施乐株式会社 图像处理装置、图像读取装置及图像形成装置
US20100092083A1 (en) * 2008-10-10 2010-04-15 Xerox Corporation In-line linear variable filter based spectrophotometer
CN101863161A (zh) * 2009-03-25 2010-10-20 海德堡印刷机械股份公司 与角度相关的颜色值修正

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP2637004A4 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103308466A (zh) * 2013-06-04 2013-09-18 沈阳仪表科学研究院有限公司 便携式滤光片色轮型多光谱成像系统及其光谱图像处理方法
CN115908174A (zh) * 2022-11-10 2023-04-04 深圳市中达瑞和科技有限公司 相对照度矫正方法、相关系统、设备和存储介质
CN119941815A (zh) * 2025-01-09 2025-05-06 武汉纺织大学 一种基于纹理特征加权校正的纱线照相测色方法及系统

Also Published As

Publication number Publication date
US9417132B2 (en) 2016-08-16
CN102466520B (zh) 2014-12-17
EP2637004A4 (en) 2015-09-23
EP2637004B1 (en) 2018-01-24
CN102466520A (zh) 2012-05-23
US20130293702A1 (en) 2013-11-07
EP2637004A1 (en) 2013-09-11

Similar Documents

Publication Publication Date Title
WO2012058977A1 (zh) 多光谱成像颜色测量系统及其成像信号处理方法
US8634014B2 (en) Imaging device analysis systems and imaging device analysis methods
US10168215B2 (en) Color measurement apparatus and color information processing apparatus
US20110176029A1 (en) Multispectral and Colorimetric Imaging System
US20030160970A1 (en) Method and apparatus for high resolution 3D scanning
JPH11331508A (ja) スキャナ及びそのスペクトル較正方法並びにスキャナを用いたイメ―ジ走査方法
JP2002303548A (ja) 平坦な測定対象物の光電測定用装置
JP2010271246A (ja) 色彩輝度測定装置及び色彩輝度測定方法
US8854707B2 (en) Imaging device analysis systems and imaging device analysis methods
JP2002320234A (ja) 二次元測定対象物をピクセル毎に光電測定するための装置
JP6606231B2 (ja) 色画像を生成するカメラおよび方法
JP2021113744A (ja) 撮像システム
US8587849B2 (en) Imaging systems, imaging device analysis systems, imaging device analysis methods, and light beam emission methods
US8334908B2 (en) Method and apparatus for high dynamic range image measurement
WO2015015061A1 (fr) Mesures de projection cinématographique
CN115183873A (zh) 一种色度检测系统、检测方法及设备
JP2020094876A (ja) 形状情報取得装置および形状情報取得方法、プログラム並びに記憶媒体
JP2009236785A (ja) 色高忠実カメラ自動測色装置
KR20220116758A (ko) 분광복사휘도 및 이미지 측정 장치 및 측정 방법
WO2018142586A1 (ja) 投射型映像表示装置
JP6997593B2 (ja) 光学特性測定方法および光学特性測定システム
CN214951806U (zh) 一种基于多光谱成像的测色装置
CN100543572C (zh) 成像系统,成像器件分析系统,成像器件分析方法,和光束发射方法
JP7491084B2 (ja) 校正装置、校正方法、校正プログラム、分光カメラ、及び情報処理装置
Gruber et al. Description and evaluation of the high quality photogrammetric scanner UltraScan 5000

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: 11837503

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2011837503

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 13882508

Country of ref document: US