WO2015021307A2 - Procédé et appareil d'évaluation de la couleur dans une image - Google Patents

Procédé et appareil d'évaluation de la couleur dans une image Download PDF

Info

Publication number
WO2015021307A2
WO2015021307A2 PCT/US2014/050194 US2014050194W WO2015021307A2 WO 2015021307 A2 WO2015021307 A2 WO 2015021307A2 US 2014050194 W US2014050194 W US 2014050194W WO 2015021307 A2 WO2015021307 A2 WO 2015021307A2
Authority
WO
WIPO (PCT)
Prior art keywords
swatch
color
sample
colors
values
Prior art date
Application number
PCT/US2014/050194
Other languages
English (en)
Other versions
WO2015021307A3 (fr
Inventor
Niraj Agarwal
Hong Wei
Original Assignee
Datacolor Holding Ag
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
Priority claimed from US13/962,495 external-priority patent/US9076068B2/en
Application filed by Datacolor Holding Ag filed Critical Datacolor Holding Ag
Publication of WO2015021307A2 publication Critical patent/WO2015021307A2/fr
Publication of WO2015021307A3 publication Critical patent/WO2015021307A3/fr

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02—Details
    • G01J3/0267—Sample holders for colorimetry
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/46—Measurement of colour; Colour measuring devices, e.g. colorimeters
    • G01J3/463—Colour matching
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/46—Measurement of colour; Colour measuring devices, e.g. colorimeters
    • G01J3/52—Measurement of colour; Colour measuring devices, e.g. colorimeters using colour charts
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/56—Extraction of image or video features relating to colour

Definitions

  • the present invention generally relates to the field of imaging, and more specifically relates to color measurement.
  • Color measurement systems help to improve operational efficiency and product quality in supply chains.
  • color approval offices for the global apparel supply chain apparel mills and dye houses, paint stores, textile printing shops, carpet manufacturers, manufacturers of wood panels, tiles, vinyl sheets, and laminates, and other industries relying on the digital color workflow require accurate color evaluation and visualization.
  • spectrophotometers are used to measure product color for solid color areas. These measurements are compared to reference standards to verify product color consistency. In some cases, the measurement and comparison data is transmitted to the buyer for remote color approval. However, some samples (e.g., printed or textured samples) cannot present a large enough area of uniform color to the spectrophotometer viewing port to allow for proper measurement.
  • a method for evaluating a color of a sample includes acquiring a color calibrated swatch of the sample, the color calibrated swatch comprising a plurality of pixels, and comparing all pixels that are of a first color in a swatch of a standard to all of the plurality of pixels that are of a second color, wherein the second color is a color in the swatch of the sample that is most similar to the first color in the swatch of the standard.
  • Another method for performing a color comparison between a first swatch comprising a first plurality of pixels and a second swatch comprising a second plurality of pixels includes computing a first set of weighted mean RGB values for the first swatch, computing a second set of weighted mean RGB values for the second swatch, and calculating a color difference between the first swatch and the second swatch, based on the first set of weighted mean RGB values and the second set of weighted mean RGB values.
  • FIG. 1 is a schematic diagram illustrating one embodiment of a system for capturing color-accurate images, according to the present invention
  • FIG. 2 is a flow diagram illustrating one embodiment of a method for evaluating a color sample, according to one embodiment of the present invention
  • FIG. 3 is a flow diagram illustrating one embodiment of a method for evaluating a color difference between a first image and a second image, according to the present invention
  • FIG. 4 is a diagram illustrating an exemplary 5 x 5 quantized image and corresponding color palette
  • FIG. 5 is a diagram illustrating one embodiment of an array produced by a step of the method illustrated in FIG. 3;
  • FIG. 6 is a high-level block diagram of the present invention that is implemented using a general purpose computing device.
  • the present invention is a method and apparatus for evaluating color in an image.
  • Embodiments of the invention perform imaging under multiple light sources that emulate International Commission on Illumination (CIE) reference illuminants in order to capture the color of a sample that cannot be measured using a spectrophotometer. Images taken in this way can be used to facilitate color difference evaluation for monochromatic, textured, and multi-colored samples. For example, a captured image of a batch sample may be compared to an image of a standard sample to verify a match. Parts of the image of a batch sample selected by the user may also be compared to selected parts of a standard sample.
  • the images and associated data can also be shared over a network and integrated in a color approval workflow.
  • FIG. 1 is a schematic diagram illustrating one embodiment of a system 100 for capturing color-accurate images, according to the present invention.
  • the main components of the system 100 include an imaging booth 102, a first computing device 104, and a second computing device 106.
  • the imaging booth 102 comprises a light box or housing 108 within which a plurality of light sources 110i-110 n (hereinafter collectively referred to as "light sources 110") are contained.
  • the imaging booth 102 houses a sample positioning mechanism such as a sample tray or pedestal 112 designed to support a sample to be measured (e.g., a fabric swatch) and a color calibration strip (e.g., a strip including a plurality of color swatches of known spectral reflectance) and/or a gray card.
  • a sample to be measured e.g., a fabric swatch
  • a color calibration strip e.g., a strip including a plurality of color swatches of known spectral reflectance
  • reference color data for the colors displayed on the calibration strip is pre-measured (e.g., using a spectrophotometer).
  • the calibration strip is integrated with the pedestal 112.
  • the light sources 110 each provide an emulation of a different type of reference illuminant. These types of reference illuminants may include, for example, daylight (e.g., emulating CIE standard "D65”), tungsten light (e.g., emulating CIE standard "Illuminant A”), or fluorescent light (e.g. , emulating CIE standard "F11").
  • the light sources 110 are positioned such that the illumination emitted by the light sources 110 is incident upon the pedestal 112, and more specifically upon the sample and calibration strip supported on the pedestal 112.
  • the imaging booth 102 additionally comprises a digital camera 114 that is positioned outside the housing 108. Specifically, the digital camera 114 is positioned in front of the housing 108, which allows the digital camera 114 to view and capture images of the sample housed within the housing 108. In one embodiment, the digital camera 114 is held in position by a bracket (not shown) affixed to the exterior of the housing 108, from which the digital camera 114 is removable. In alternative embodiments, the imaging booth 102 may be configured as an imaging unit. For example, the imaging booth 102 and all of its components may be scaled to fit within a handheld device that has an open port for positioning the sample. In such a case, the light sources 110 and digital camera 114 may also be part of the handheld device.
  • the first computing device 104 is communicatively coupled to the imaging booth 102 and to the digital camera 114.
  • the first computing device 104 is configured to acquire and process images captured by the digital camera 114.
  • the first computing device 104 converts the camera RGB values of the captured images to tristimulus XYZ values, as discussed in further detail below.
  • the first computing device 104 stores processed images and associated data (and possibly meta data relating to the sample) on a server 116.
  • the server may also store data relating to a standard sample.
  • the first computing device 104 comprises a general purpose computing device or a special purpose computing device, such as a personal computer, a laptop computer, a tablet computer, a smart phone, or the like.
  • the second computing device 106 is a remote computer that is communicatively coupled to the server 116.
  • the second computing device 106 is configured to retrieve and display the processed images for viewing and color evaluation (and thus comprises a display and/or printer).
  • Color evaluation using the second computing device 106 may comprise visual evaluation and/or numerical evaluation that simulates visual evaluation, as discussed below.
  • the second computing device 106 comprises a general purpose computing device or a special purpose computing device, such as a personal computer, a laptop computer, a tablet computer, a smart phone, or the like.
  • the second computing device 106 is coupled to a screen color calibration device 118 such as a screen colorimeter.
  • the screen color calibration device 118 calibrates the colors displayed by the second computing device 106 so that visual color evaluation performed using the second computing device 106 is reliable.
  • another screen color calibration device may be coupled to the first computing device 106.
  • a single computing device replaces the first computing device 104 and the second computing device 106.
  • both processing and display of the images occur on a single computing device.
  • color evaluation and image color difference computation may also occur on this single computing device.
  • the server 116 may not be necessary.
  • three different computing devices are used.
  • a first computing device processes the images as described above and a second computing device displays the images as described above.
  • a third computing device then performs computations that judge the similarity of the colors.
  • FIG. 2 is a flow diagram illustrating one embodiment of a method 200 for evaluating a color sample, according to one embodiment of the present invention.
  • the method 200 may be implemented, for example, in conjunction with the system 100 illustrated in FIG. 1.
  • the method 200 is not limited to implementation with the configuration illustrated in FIG. 1, however, and may be implemented in conjunction with systems whose configurations differ from the system 100.
  • the method 200 is implemented in step 202 and proceeds to step 204, where the settings of the system 100 are prepared for image capture.
  • the sample to be evaluated and the calibration strip are positioned on the pedestal 112 of the imaging booth 102.
  • One or more of the light sources 110 also is selected to illuminate the sample. In one embodiment, a sequence by which each selected light source 110 illuminates the sample one-by-one is specified.
  • the digital camera 114 is positioned such that the sample and the calibration strip are within its field of view, and the gray card is used to calibrate the digital camera 114 for light intensity non-uniformity in the imaging plane.
  • the digital camera 114 captures one or more images of the sample and the calibration strip. Any given one of the captured images will display both the sample and the calibration strip.
  • multiple images under different reference illuminants may be captured. Imaging under multiple reference illuminants allows color capture from samples that cannot be measured using a spectrophotometer (as a spectrophotometer has too big an aperture to measure from small elements of a multicolored sample).
  • step 208 the first computing device 104 corrects the captured images for light non-uniformity, in accordance with the gray card. Then, in step 210, the first computing device calibrates the color of the captured images, in accordance with the calibration strip, to produce color calibrated output images. In one embodiment, the first computing device 104 operates on the captured images in the RAW image file format.
  • step 212 the color calibrated output images and/or meta data related to the sample is stored.
  • the output images and data are stored locally on the first computing device 104.
  • the output images and data are delivered by the first computing device 104 over a network to the server 1 16 for remote storage.
  • the color calibrated output images are visually reviewed for approval.
  • review of the color calibrated output images involves retrieving the output images over the network and displaying the output images by the second computing device 106.
  • the output images may be displayed on the first computing device 104.
  • images of multiple samples may be viewed at once on the display.
  • visual review is based on the display RGB values of the color calibrated output images (where the display RGB values are calculated from the XYZ tristimulus values).
  • Visual review is facilitated by calibrating the screen color of the display using an International Color Consortium (ICC) profile created by the screen color calibration device 1 18.
  • ICC International Color Consortium
  • the color calibrated output images are numerically reviewed.
  • the second computing device 106 (or first computing device 104) executes a program that calculates the difference in color appearance between a pair of swatches represented in an image (e.g., of a standard sample and of a batch sample).
  • the swatches e.g. , the standard sample and the batch sample
  • a "swatch" refers to a set of pixels from a part of an image that contains either a batch sample or a standard sample.
  • all pixels of a given color in one of the swatches are compared to all pixels of the most similar color in the other swatch (e.g., of the standard sample).
  • This comparison is done after the process of converting the camera RGB values of the captured images to XYZ tristimulus values.
  • display and review on the second computing device 106 allows for review and approval of samples based on remotely captured images.
  • step 218 The method 200 terminates in step 218.
  • the method 200 allows remote parties to verify color consistency between a standard sample and a set of batch samples, based on intra- or inter- image comparison (depending on whether the standard and the batch samples are represented in the same or different images).
  • the image comparison visually and numerically compares the colors of the standard sample and the batch samples. In this way, color consistency can be assured without the need to ship physical samples, and the production history for the color products can be archived for future reference.
  • the first computing device 104 corrects the captured images for light non-uniformity, using a gray card. This process is well-known and involves scaling each pixel intensity to the mean pixel intensity in every image plane (R, G, B for each camera illuminant), image-by-image.
  • P be any one of the camera R, G, B image channels for any of the plurality of camera illuminants.
  • P(i, j) be the ij pixel of the P image in the presence of test colors
  • P 0 (i, j) be the ij pixel of the P image in the presence of the gray card. Then, for each of the nine P channels, the following steps are performed.
  • the image Po(i, j) and its mean P me an is acquired for all i,j.
  • the test color image P(i, j) is acquired.
  • the present invention allows images of samples to be color calibrated based on multiple images of the samples captured under different reference illumination conditions.
  • this calibration involves estimating the XYZ tristimulus values of a sample's reflectance under three illuminants, from the RGB camera values and from the RGB camera values of a simultaneously-imaged chart of calibration samples. These calibration samples together comprise a calibration chart. These operations are performed, for example, in accordance with step 216 of the method 200.
  • the tristimulus values of the reflectance under the three reference illuminants are estimated from the reflectance's measured camera RGB values under the same three illuminants (column 9-vector d) in conjunction with a number K of simultaneously acquired calibration samples whose pre-measured reflectance spectra are used to compute tristimulus values that are stored in a database.
  • the calibration values in this case comprise a 9 x K matrix D of camera values from the calibration chart and a 9 x K matrix R T of tristimulus values from the same calibration chart, where K is the number of reflectances.
  • K the number of reflectances.
  • the matrix D from the calibration chart is measured by the camera under the same illuminants, and at the same times, as is an image of test reflectances, and the matrix R T is pre-computed from the reflectance values pre-measured from the calibration chart, in conjunction with color matching functions and three reference-illuminant spectra.
  • a 9 x 9 matrix M and an offset vector b are computed so as to estimate the tristimulus 9-vector r of any test reflectance.
  • the matrix M and offset vector b map (as closely as possible) all of the database camera value 9-vectors to corresponding tristimulus 9-vectors as follows:
  • D' is the transpose of D.
  • a 3 x K matrix stores the data for each illuminant handled separately.
  • the 3 x 3 calibration matrix M is calculated separately for each illuminant, and each image is corrected using the corresponding M matrix.
  • R T is a 9 x 24 matrix, where R T (1 ,1) is the R value of reflectance 1 under camera illuminant 1; R T (1 ,2) is the R value of reflectance 2 under camera illuminant 1 ; RT(2,1) is the G value of reflectance 1 under camera illuminant 1 ; and RT(2,2) is the G value of reflectance 2 under camera illuminant 1.
  • the regions to be compared are excerpted from the color-calibrated batch sample image and from the color-calibrated standard sample image (which, as described above, may be the same image). All of the images at this point are display-RGB images (converted from XYZ images) and have the same destination illuminant. Optionally, the images are quantized into a smaller number of colors (e.g., 256 colors). Quantization may be performed using, for example, the methods described Wu in "Efficient Statistical Computations for Optimal Color Quantization” (Graphics Gems, vol. II, p.126-133).
  • the standard and batch swatches are quantized separately, but it is also possible to quantize the standard and the batch swatches to a common palette.
  • the camera RGB values for the whole image or images from which the swatches were derived were converted to tristimulus XYZ values under each light using the (10 x 9) M matrix, described above.
  • the quantization can be performed using display RGB values. However, in this case, if subsequent processing steps use the tristimulus XYZ values, then the display RGB values (which were derived from the tristimulus XYZ values) must be transformed back into tristimulus XYZ values. Techniques for performing the display RGB-to-tristimulus XYZ transformation are known to those of skill in the art and are not discussed in detail here.
  • the color difference relative to every color in the batch swatch is computed.
  • the color difference for a color pair is computed as ⁇ , which describes the distance between two colors in color space.
  • ⁇ for the closest color in the batch swatch is weighted by the pixel frequency of the color in the batch swatch.
  • the mean standard RGB values are also weighted by the pixel frequency of the color in the standard swatch.
  • the weighted sum over all colors in the standard swatch provides a color difference measure that correlates with a visual comparison between the standard swatch and the batch swatch.
  • the same batch swatch color may be mapped to multiple standard swatch colors.
  • the percent of standard sample colors (weighted by total pixel count for the standard sample color) that can be matched to any color in the batch samples (e.g., within a tolerance of ⁇ ) is identified.
  • the percent of the batch colors that are matched to the standard colors is also identified. These numbers are correlated, but are also individual outputs.
  • the present invention evaluates the color difference between a first swatch (e.g., depicting a standard sample) and a second swatch (e.g., depicting batch sample) by comparing all of the pixels of one color in the first swatch with all of the pixels of the most similar color in the second swatch.
  • two color matching indices are calculated to represent the color different between the first swatch and the second swatch.
  • two images are generated for each light source: one XYZ image (e.g., sixteen bits, floating, with an .exr file extension) and one display RGB image (e.g., eight bits, integer, with a .png file extension).
  • one XYZ image e.g., sixteen bits, floating, with an .exr file extension
  • one display RGB image e.g., eight bits, integer, with a .png file extension
  • an overall match index is referred to as "Mean ⁇ " or " ⁇ of Mean,” which is the color difference (i.e., AECMC) between the arithmetic mean of the colors of the first swatch and the arithmetic mean of the colors of the second swatch.
  • the XYZ images are used in this calculation.
  • the CIE tristimulus vector of the first swatch is obtained by
  • the vectors can be used to calculate the overall match index ⁇ (for example AECMC)-
  • a visual match index is referred to as "Visual ⁇ " or "Best Match ⁇ .”
  • the calculation of the visual match index essentially "looks for the standard in the batch.” Rather than averaging all of the pixels in first and second swatches as the overall match index does, the visual match index calculation compares local colors in the first and second swatches in a manner that resembles the visual assessment performed by the human eye.
  • the visual match index calculation is in one embodiment based on the display RGB images (e.g., .png files). Basing the visual match index on the display RGB images is intended to simplify processing (which technically is possible - but more difficult - when based on the XYZ images).
  • FIG. 3 is a flow diagram illustrating one embodiment of a method 300 for evaluating a color difference between a first swatch [e.g., depicting a standard sample) and a second swatch (e.g., depicting a batch sample), according to the present invention.
  • the method 300 describes how to calculate the visual match index described above.
  • the method 300 begins in step 302.
  • the colors of the first and second swatches are both quantized to reduce the number of distinct colors.
  • the swatches are assumed to be quantized to 256 colors for the following explanation.
  • the direct outputs of the quantization process comprise a color palette and a two-dimensional image.
  • the color palette comprises a 256 x 3 array in which each row represents a color and each column represents an R, G, or B value for each color.
  • the two-dimensional image is indexed such that each pixel in the two-dimensional image contains only a corresponding color index (i.e., row number) in the array. Counting the pixels in the two- dimensional image yields a 256 x 1 array containing a histogram of the total number of pixels for each indexed color.
  • FIG. 4 is a diagram illustrating an exemplary 5 x 5 quantized image and corresponding color palette.
  • Each square in the image represents a single pixel with a coded color.
  • the color palette comprises a 4 x 3 array in which the R, G, and B values of the four indexed colors are contained.
  • a corresponding two-dimensional indexed image indicates the color index (e.g. , zero through three) contained in each pixel. From the indexed image, the frequency of each indexed color can be calculated and used to produced a 4 x 1 array as shown.
  • step 304 produces four arrays: (1) a color palette of the first swatch (e.g., a 256 x 3 array); (2) a color palette of the second swatch (e.g., a 256 x 3 array); (3) a color frequency array of the indexed colors in the first swatch (e.g., a 256 x 1 array); and (4) a color frequency array of the indexed colors in the second swatch (e.g., a 256 x 1 array).
  • a color palette of the first swatch e.g., a 256 x 3 array
  • a color palette of the second swatch e.g., a 256 x 3 array
  • a color frequency array of the indexed colors in the first swatch e.g., a 256 x 1 array
  • a color frequency array of the indexed colors in the second swatch e.g., a 256 x 1 array
  • step 306 the XYZ values of each indexed color in the first and second swatches are computed.
  • this involves converting the quantized display RGB values in each row of the color palettes of the first and second swatches into XYZ values.
  • these display RGB values are linearized with respect to energy.
  • RGB linearization may be accomplished as follows: where v represents the linearized r, g, or b values ve ⁇ r, g, b ⁇ and V represents the original R, G, or B values Ve ⁇ R, G, B ⁇ .
  • the XYZ values are further chromatically adapted to the XYZ values depending on the light source white point.
  • the chromatic adaptation may be a linear Bradford transformation, achieved by color management techniques that are known in the art.
  • step 306 produces four arrays: (1) the XYZ values of all of the colors in the quantized first swatch (e.g., a 256 x 3 array); (2) the XYZ values of all of the colors in the quantized second swatch (e.g., a 256 x 3 array); (3) the total number of pixels in the first swatch (e.g., an integer); and (4) the total number of pixels in the second swatch (e.g., an integer).
  • step 308 for each color in the first swatch, the closest corresponding color (i.e., the color having the smallest color difference) is found in the second swatch. More specifically, for each quantized color [X,Y,Z] k ,i in the first swatch (where [X,Y,Z] k i is the quantized color in the k th row of the array of the XYZ values of all of the colors in the quantized first swatch), the one quantized color [X,Y,Z] m 2 in the second swatch (which is, for example, in the m th row of the array of the XYZ values of all of the colors in the quantized second swatch) that has the smallest color difference (for example, AE C MC) is found. Each pair of quantized colors [k, m] is recorded. It is possible that one quantized color in the second swatch may be matched to multiple quantized colors in the first swatch.
  • step 308 produces one array (e.g., a 256 x 2 array).
  • FIG. 5 is a diagram illustrating one embodiment of an array produced by step 308 of the method 300.
  • the first column of the array represents the index of the colors in the first swatch
  • the second column of the array represents the index of the color in the second swatch that has the smallest color difference relative to the corresponding color from the first swatch (i.e. , the color whose index is listed in the same row of the array).
  • the weighted mean RGB values of the first and second display RGB images corresponding to the first and second swatches are calculated.
  • the mean RGB of the first swatch is calculated by averaging the colors at the first column of the array produced in step 308, after each row of the color palette is weighted by the total count of the corresponding color recorded in the corresponding row of the color frequency array of the indexed colors in the first swatch. Mathematically, this may be expressed as:
  • Colori is the color palette of the first swatch
  • Frequencyi is the color frequency array of the indexed colors in the first swatch
  • Pixelsi is the total number of pixels in the first swatch.
  • EQN. 9 assumes that the first swatch has been quantized to 256 colors; however, alternative quantization schemes are easily accommodated by substitution. EQN. 9 also ignores those colors that have very small frequencies.
  • the cutoff threshold for the frequency is defined based on the specific application and may range from 0.1% to 0.2%.
  • RGB 2 R 2 G 2 B 2
  • step 310 produces two outputs: (1) a vector (e.g., 1 x 3) containing the weighted mean RGB values of the first swatch; and (2) a vector (e.g., 1 x 3) containing the weighted mean RGB values of the second swatch.
  • a vector e.g., 1 x 3 containing the weighted mean RGB values of the first swatch.
  • step 312 the visual match index (Visual ⁇ ) is calculated for the first and second swatches. In one embodiment, this involves first converting the weighted mean RGB values calculated in step 310 into XYZ values with corresponding chromatic adaptation. This may be accomplished in the same manner described above in connection with step 306.
  • step 312 produces one output: a visual match index that may be output (e.g., displayed or sent) to a user.
  • step 314 two lists of matched colors are produced, based on a predefined color difference threshold (AE t0 ierance)-
  • the first list includes all of the colors in the batch swatch that are close to each color in the standard swatch.
  • the second list includes all of the colors in the standard swatch that are close to each color in the batch swatch.
  • the "closeness" of the colors is gated by the color difference threshold AE t0 ierance-
  • the predefined threshold (AE t0 ierance) is either hardcoded or user- definable. In an exemplary embodiment, the predefined threshold is 2.0.
  • step 314 produces two lists: (1) a list of colors in the second swatch that match each color in the first swatch; and (2) a list of colors in the first swatch that match each color in the second swatch.
  • step 316 the color percentages are calculated.
  • step 316 calculates the percentage of the pixels in the second swatch whose colors are present in the first swatch, and vice versa.
  • the number of pixels in the first swatch that have the colors listed in the list of colors in the first swatch that match each color in the second swatch may be summed. This sum is then divided by the total number of pixels in the first swatch. This can be expressed mathematically as follows:
  • i in list is the colors listed in the list of colors in the first swatch that match each color in the second swatch (e.g., as produced in step 314).
  • the number of pixels in the second swatch that have the colors listed in the list of colors in the second swatch that match each color in the first swatch may be summed. This sum is then divided by the total number of pixels in the second swatch. This can be expressed mathematically as follows:
  • Match 2 100 x ⁇ . 3 ⁇ 4 Frequency 2 ( ) / Pixels 2 (EQN. 12) where "i in list” is the colors listed in the list of colors in the second swatch that match each color in the first swatch (e.g., as produced in step 314).
  • step 316 produces two values: (1) the percentage of the colors in the first swatch that are present in the second swatch; and (2) the percentage of the colors in the second swatch that are present in the first swatch. These percentages may be output (e.g., displayed or sent) to a user.
  • the method 300 ends in step 318.
  • the steps of the method 300 may also be performed in several alternative orders. For instance, referring to the left-hand side of FIG. 3, the method 300 may terminate immediately after step 312 is performed (i.e., bypassing steps 314-316, discussed above). Alternatively, referring to the right-hand side of FIG. 3, step 306 may be followed directly by step 314. Once step 314 is performed, the method 300 may proceed to steps 316, 308, 310, and 312 before ending in step 318. As a further alternative, the method 300 may terminate immediately after step 316 is performed (i.e., bypassing steps 308-312). Thus, some of the steps of the method 300 may be performed in different orders, and some of the steps may also be considered optional. The same reference numerals have been used for the corresponding steps on the left- and right-hand sides of FIG. 3 in order to make clear that the same operations are being performed, simply in a different order.
  • the values of the overall match index (Mean ⁇ ) and the visual match index (Visual ⁇ ) are not correlated.
  • the visual match index is not necessarily smaller than the overall match index.
  • the overall match index is a mathematical averaging view of the color difference, whereas the visual match index is more sensitive to local color variations.
  • FIG. 6 is a high-level block diagram of the present invention that is implemented using a general purpose computing device 600.
  • a general purpose computing device 600 comprises a processor 602, a memory 604, an evaluation module 605 and various input/output (I/O) devices 606 such as a display, a keyboard, a mouse, a stylus, a wireless network access card, a colorimeter, and the like.
  • I/O devices such as a display, a keyboard, a mouse, a stylus, a wireless network access card, a colorimeter, and the like.
  • at least one I/O device is a storage device (e.g., a disk drive, an optical disk drive, a floppy disk drive).
  • the evaluation module 605 can be implemented as a physical device or subsystem that is coupled to a processor through a communication channel.
  • the evaluation module 605 can be represented by one or more software applications (or even a combination of software and hardware, e.g., using Application Specific Integrated Circuits (ASIC)), where the software is loaded from a storage medium (e.g., I/O devices 606) and operated by the processor 602 in the memory 604 of the general purpose computing device 600.
  • a storage medium e.g., I/O devices 606
  • the evaluation module 605 for evaluating the color of an image as described herein with reference to the preceding Figures can be stored on a computer readable storage device (e.g., RAM, magnetic or optical drive or diskette, and the like).
  • one or more steps of the methods described herein may include a storing, displaying and/or outputting step as required for a particular application.
  • any data, records, fields, and/or intermediate results discussed in the methods can be stored, displayed, and/or outputted to another device as required for a particular application.
  • steps or blocks in the accompanying Figures that recite a determining operation or involve a decision do not necessarily require that both branches of the determining operation be practiced. In other words, one of the branches of the determining operation can be deemed as an optional step.
  • top, side, bottom, front, back, and the like are relative or positional terms and are used with respect to the exemplary embodiments illustrated in the Figures, and as such these terms may be interchangeable.

Landscapes

  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Spectrometry And Color Measurement (AREA)

Abstract

L'invention concerne un procédé d'évaluation d'une couleur d'un échantillon qui comprend l'acquisition d'un échantillon de couleur étalonné de l'échantillon, l'échantillon de couleur étalonné comportant une pluralité de pixels, et la comparaison, de tous les pixels qui sont d'une première couleur dans un échantillon d'un étalon, à l'ensemble de la pluralité de pixels qui sont d'une seconde couleur, la seconde couleur étant une couleur dans l'échantillon de couleur de l'échantillon qui est la plus proche de la première couleur dans l'échantillon de couleur de l'étalon.
PCT/US2014/050194 2013-08-08 2014-08-07 Procédé et appareil d'évaluation de la couleur dans une image WO2015021307A2 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US13/962,495 US9076068B2 (en) 2010-10-04 2013-08-08 Method and apparatus for evaluating color in an image
US13/962,495 2013-08-08

Publications (2)

Publication Number Publication Date
WO2015021307A2 true WO2015021307A2 (fr) 2015-02-12
WO2015021307A3 WO2015021307A3 (fr) 2015-04-09

Family

ID=52462044

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2014/050194 WO2015021307A2 (fr) 2013-08-08 2014-08-07 Procédé et appareil d'évaluation de la couleur dans une image

Country Status (1)

Country Link
WO (1) WO2015021307A2 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107346429A (zh) * 2017-05-24 2017-11-14 上海电机学院 一种多晶电池片图像的颜色自动识别与分类方法
WO2017205119A1 (fr) * 2016-05-27 2017-11-30 Microsoft Technology Licensing, Llc Palette d'étalonnage pour caméra thermique

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3413589B2 (ja) * 1998-03-09 2003-06-03 富士写真フイルム株式会社 画像処理方法および装置
US7162077B2 (en) * 2001-10-19 2007-01-09 Sharp Laboratories Of America, Inc. Palette-based image compression method, system and data file
US20100104178A1 (en) * 2008-10-23 2010-04-29 Daniel Tamburrino Methods and Systems for Demosaicing
CN102687007B (zh) * 2009-09-18 2015-07-22 罗格斯州立大学 利用分层标准化切割的高处理量生物标志物分割
US8532371B2 (en) * 2010-10-04 2013-09-10 Datacolor Holding Ag Method and apparatus for evaluating color in an image

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017205119A1 (fr) * 2016-05-27 2017-11-30 Microsoft Technology Licensing, Llc Palette d'étalonnage pour caméra thermique
CN107346429A (zh) * 2017-05-24 2017-11-14 上海电机学院 一种多晶电池片图像的颜色自动识别与分类方法

Also Published As

Publication number Publication date
WO2015021307A3 (fr) 2015-04-09

Similar Documents

Publication Publication Date Title
US9076068B2 (en) Method and apparatus for evaluating color in an image
US8532371B2 (en) Method and apparatus for evaluating color in an image
CN108020519B (zh) 一种基于颜色恒常性的虚拟多光源光谱重建方法
US7940393B2 (en) Method and system for approximating the spectrum of a plurality of color samples
CN109963133A (zh) 色彩校正系统和方法
CN110926609A (zh) 一种基于样本特征匹配的光谱重建方法
CN112488997B (zh) 基于特征插值的古代绘画印刷品颜色复现检测和评价方法
WO2015021307A2 (fr) Procédé et appareil d'évaluation de la couleur dans une image
US10194035B2 (en) Imager calibration via modeled responses to importance-weighted color sample data
JP4174707B2 (ja) 分光測定システム、色再現システム
EP2672719A2 (fr) Étalonnage de la couleur d'un dispositif de capture d'images de manière à s'adapter à la scène à capturer
CN110926608A (zh) 一种基于光源筛选的光谱重建方法
Nieves et al. Recovering fluorescent spectra with an RGB digital camera and color filters using different matrix factorizations
Zhu et al. Color calibration for colorized vision system with digital sensor and LED array illuminator
CN107079072B (zh) 成像器的交叉校准
Colantoni et al. A color management process for real time color reconstruction of multispectral images
Zhao et al. Methods of spectral reflectance reconstruction for a Sinarback 54 digital camera
Shrestha Multispectral imaging: Fast acquisition, capability extension, and quality evaluation
Carnevali et al. Colourimetric Calibration for Photography, Photogrammetry, and Photomodelling Within Architectural Survey
JP4189493B2 (ja) 色彩コンフォートメータ
Grossberg et al. Estimating true color imagery for GOES-R
KR101366163B1 (ko) 색지각 조건을 대응한 색재현 성능 최적화 하이브리드 방법 및 시스템
ES2953945T3 (es) Visualización de tinción de madera
Eckart et al. Deriving absolute color from 6-band visible WorldView-2 and-3 imagery using reflectance spectra reconstruction via principal component analysis and conversion to L* a* b
Pointer et al. Food colour appearance judged using images on a computer display

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: 14834009

Country of ref document: EP

Kind code of ref document: A2

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14834009

Country of ref document: EP

Kind code of ref document: A2