Printing digital original image evaluation method and system based on memory color
By calculating the color, clarity and order index, combined with the memory color image data set, the problem that image quality evaluation in the prior art is difficult to accurately reflect the quality of the original image of the printed digital, and objective evaluation of the printed materials is achieved, reducing manpower and material resources and improve the consistency and accuracy of evaluation.
Patent Information
- Application Number
- CN202411010793.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-12
AI Technical Summary
Existing image quality evaluation methods are difficult to accurately reflect the quality of printed digital original images, especially in terms of color tone, clarity and memory colors, and objective evaluation indicators cannot fully reflect the visual perception effect of the human eye.
The original image evaluation method based on memory color is used to evaluate the color index, clarity index and order index, combined with the memory color image data set, and the model (such as Y=γ1X1′+γ2X2′+γ3X3′) is used. Reduce the consumption of manpower and material resources and improve the objectivity and consistency of evaluation.
Effectively evaluate the color error and noise interference of printed digital originals during transmission, storage, compression and other processing, which improves the objective reflection of image quality, and the evaluation score is consistent with the visual perception characteristics of the human eye, reducing the cost loss of printed materials.
Smart Images

Figure CN118941536A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of image data processing, and in particular to a method and system for evaluating printed digital original images based on memory colors. Background Art
[0002] Memory color is the long-term and stable color of common objects formed in people's minds through visual perception. For example, skin color, grass, sky color, etc. are all memory colors. The accurate reproduction of memory color plays a key role in evaluating the image quality of familiar objects in actual scenes. The color appearance attributes presented by people are compared with the long-term memory color of the object in people's impression. If the color appearance attributes are more consistent with the memory color, it is more satisfactory.
[0003] The purpose of printing and reproduction is to reproduce the color gradation of the original. Nowadays, more and more images are digital, but the transmission between different devices involves the acquisition, compression, storage and transmission of images. Hardware conditions and environmental factors may cause Gaussian noise, block effect, ringing, distortion and blur, which will affect the accuracy of image printing. The colors in the image will also be different from the colors in memory. There may be overall color cast, low-key color cast, high-key color cast, and different colors of high and low-key color casts. The printed product will also affect people's visual perception.
[0004] Existing image evaluation methods also involve a variety of distortions, but there is no index based on the subjective and objective comprehensive evaluation of memory color. Using some characteristics of the human visual system and modeling based on certain assumptions, it is difficult to meet the requirements of stable and accurate image quality evaluation. In addition, the existing technology evaluates image quality based on a variety of objective evaluation indicators. However, in the process of obtaining objective evaluation indicators, the image will undergo multiple levels of processing, which often introduces a variety of distortions and cannot fully reflect the human eye's perception of image quality.
[0005] Therefore, the existing image quality evaluation methods still have much room for improvement in terms of tonality, clarity, saturation and memory color evaluation. Summary of the invention
[0006] The present invention is proposed based on the above-mentioned needs of the prior art. The technical problem to be solved by the present invention is to provide a method and system for evaluating printed digital original images based on memory colors to objectively reflect whether the printed digital original images are qualified printed manuscripts.
[0007] In order to solve the above problems, the present invention is implemented by adopting the following technical solutions:
[0008] A method for evaluating a printed digital original image based on memory color, the method comprising: obtaining a printed digital original image, the printed digital original image being an RGB image as a digital image manuscript for various types of printing; calculating a first color index, comprising: converting the printed digital original image into a LAB image; calculating the average values of L, A and B channels in the LAB image, and calculating the offset between the average value of the L channel and 50, the average value of the A channel and the average value of the B channel to determine the first color index; calculating a first clarity index, comprising: determining a gradient in a horizontal direction and a gradient in a vertical direction of the printed digital original image, calculating a gradient amplitude value of each pixel based on the gradient in the horizontal direction and the gradient in the vertical direction, and obtaining a maximum amplitude value among the gradient amplitude values as a first maximum amplitude value; extracting a central area from the printed digital original image based on the number of rows and columns of pixels contained in the printed digital original image; obtaining an average value, a second maximum amplitude value and a minimum amplitude value of the gradient amplitude values in the central area, wherein the second maximum amplitude value is the maximum gradient amplitude value in the central area; calculating a first clarity index based on the second maximum amplitude value and the minimum amplitude value The invention relates to a method for determining a first color index, a first clarity index, and a first tone index, and a first color index. The method comprises the following steps: determining a first color index, a first clarity index, and a first tone index based on the first maximum amplitude value and the gradient variability value; determining a first clarity index based on the first maximum amplitude value and the gradient variability value; calculating a first tone index, including: counting the number of occurrences of each pixel value in the printed digital original image, the maximum pixel value and the minimum pixel value in the printed digital original image, and calculating a histogram based on the number of occurrences of each pixel value; obtaining a first percentile and a third percentile of the histogram; determining a first tone index based on the difference between the third percentile and the first percentile and the ratio of the difference between the maximum pixel value and the minimum pixel value; inputting the first color index, the first clarity index and the first tone index into a first model to obtain evaluation data of the printed digital original image, wherein the expression of the first model is Y=γ1X1′+γ2X2′+γ3X3′, wherein Y represents a quality score, X1′ represents a color index to be input, X2′ represents a clarity index to be input, X3′ represents a tone index to be input, γ1 represents a first parameter, γ2 represents a second parameter, and γ3 represents a third parameter.
[0009] Optionally, the offset between the average value of the L channel and 50, the average value of the A channel and the average value of the B channel are calculated to determine a first color index, whose expression includes: Among them, S a Indicates the saturation of the color, b1 indicates the offset between the average value of channel L and 50, b2 indicates the average value of channel A, b3 indicates the average value of channel B, and X1 indicates the first color index.
[0010] Optionally, extracting the central area from the printed digital original image based on the number of rows and columns of pixels contained in the printed digital original image comprises: Wherein, m represents the number of rows of pixels contained in the printed digital original image, n represents the number of columns of pixels contained in the printed digital original image, and B represents a constant to determine the size and position of the central area; the central area is extracted based on the rows and columns between the B+1th pixel to the Bth pixel from the end in the printed digital original image.
[0011] Optionally, the gradient variability value is determined based on the ratio of the difference between the second maximum amplitude value and the minimum amplitude value to the average of the gradient amplitude values in the central area, and its expression includes: Among them, VG represents the gradient variability value, MaxG represents the second maximum amplitude value, MinG represents the minimum amplitude value, i represents the i-th row of the central area, j represents the j-th column of the central area, G(i,j) represents the value of the i-th row and j-th column of the central area, W represents the width of the central area, and H represents the height of the central area.
[0012] Optionally, a first clarity index is determined based on the first maximum amplitude value and the gradient variability value, and its expression includes: X2=MG α ×VG 1-α , wherein X2 represents the first clarity index, MG represents the first maximum amplitude value, VG represents the gradient variability value, and α represents a constant.
[0013] Optionally, an expression for calculating the first tone index includes: Among them, X3 represents the first tone index, y represents the third percentile of the histogram, x represents the first percentile of the histogram, ma represents the maximum pixel value, and mi represents the minimum pixel value.
[0014] Optionally, determining the first model includes: acquiring a memory color image data set; randomly sampling the memory color image data set, and distorting the sampled image to obtain a distorted image, wherein the distortion processing includes adjusting the saturation, contrast or Gaussian noise variance; within a preset time, subjectively scoring the sampled image and the distorted image to obtain a scoring score, and calculating a second color index, a second clarity index and a second tone index; inputting the scoring score as Y, the second color index as X1′, the second clarity index as X2′, and the second tone index as X3′ into Y=γ1X1′+γ2X2′+γ3X3′ to determine the first parameter, the second parameter and the third parameter to obtain the first model.
[0015] Optionally, the expression of the first model is: Y=0.4159×X1′+0.2874×X2′+0.0391×X3′.
[0016] A printing digital original image evaluation system based on memory color, the system comprises: an acquisition module, which acquires a printing digital original image, wherein the printing digital original image is an RGB image; a color index module, which calculates a first color index, comprising: converting the printing digital original image into a LAB image; calculating the average values of L, A and B channels in the LAB image, and calculating the offset between the average value of the L channel and 50, the average value of the A channel and the average value of the B channel, and determining the first color index; a clarity index calculation module, which calculates the first clarity index, comprising: determining the gradient in the horizontal direction and the gradient in the vertical direction of the printing digital original image, calculating the gradient amplitude value of each pixel based on the gradient in the horizontal direction and the gradient in the vertical direction, and acquiring the maximum amplitude value in the gradient amplitude values as the first maximum amplitude value; extracting a central area from the printing digital original image based on the number of rows and columns of pixels contained in the printing digital original image; acquiring the average value, the second maximum amplitude value and the minimum amplitude value of the gradient amplitude value in the central area, wherein the second maximum amplitude value is the maximum gradient amplitude value in the central area; calculating the first clarity index based on the difference between the second maximum amplitude value and the minimum amplitude value; The method comprises the following steps: calculating the first gradation index by calculating the ratio of the difference between the first maximum amplitude value and the average value of the gradient amplitude value in the central area; determining the gradient variability value; determining the first clarity index based on the first maximum amplitude value and the gradient variability value; a gradation index calculation module, calculating the first gradation index, including: counting the number of occurrences of each pixel value in the printed digital original image, the maximum pixel value and the minimum pixel value in the printed digital original image, and calculating a histogram based on the number of occurrences of each pixel value; obtaining the first percentile and the third percentile of the histogram; calculating the first gradation index based on the difference between the third percentile and the first percentile, and the maximum pixel value and the minimum pixel value. The first color index, the first clarity index and the first tone index are determined by the ratio of the difference between the minimum pixel values; the output module inputs the first color index, the first clarity index and the first tone index into the first model to obtain the evaluation data of the printed digital original image, and the expression of the first model is Y=γ1X1′+γ2X2′+γ3X3′, wherein Y represents the quality score, X1′ represents the color index to be input, X2′ represents the clarity index to be input, X3′ represents the tone index to be input, γ1 represents the first parameter, γ2 represents the second parameter, and γ3 represents the third parameter.
[0017] Optionally, determining the first model includes: acquiring a memory color image data set; randomly sampling the memory color image data set, and distorting the sampled image to obtain a distorted image, wherein the distortion processing includes adjusting the saturation, contrast or Gaussian noise variance; within a preset time, subjectively scoring the sampled image and the distorted image to obtain a scoring score, and calculating a second color index, a second clarity index and a second tone index; inputting the scoring score as Y, the second color index as X1′, the second clarity index as X2′, and the second tone index as X3′ into Y=γ1X1′+γ2X2′+γ3X3′ to determine the first parameter, the second parameter and the third parameter to obtain the first model.
[0018] Compared with the prior art, the present invention provides a method for evaluating printed digital original images based on memory colors, which objectively reflects whether the printed digital original image is a qualified printed manuscript, reduces the consumption of manpower and material resources, and its evaluation score is consistent with the visual perception characteristics of the human eye for the image. It integrates color, clarity, gradation, and subjective consistency, and combines with the memory color image set to effectively evaluate the color error, noise interference, and gradation change caused by the distortion of image quality during the acquisition, transmission, storage, compression, decompression, and editing of the printed digital manuscript image. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this specification. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.
[0020] Figure 1 is a flow chart of a printed digital original image evaluation method based on memory color provided in this embodiment;
[0021] Figure 2 It is a schematic diagram of the data transmission process of a printing digital original image evaluation method based on memory color provided in this embodiment. DETAILED DESCRIPTION
[0022] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0023] To facilitate understanding of the embodiments of the present invention, the following will further explain and illustrate specific embodiments in conjunction with the accompanying drawings. The embodiments do not constitute a limitation on the protection scope of the present invention.
[0024] Example 1
[0025] In the prior art, the existing image quality evaluation is based on the digital image itself, and the evaluation angle is based on the evaluation of the parameters of the image itself, which has nothing to do with printing. In the printing environment, some unique evaluation methods are needed to better evaluate the color compression, gradation compression, and clarity of the original manuscript of the printed product. On the other hand, unlike traditional digital images, the subjective evaluation criteria for the color quality of printed products are to reproduce the truth, nature, and coordination, which needs to be approved by customers, and the evaluation methods adopted by customers are mostly subjective evaluations based on color. Therefore, this embodiment uses a subjective evaluation method to simulate the customer's evaluation, establish a memory color data set for the color distortion of memory colors, objectively evaluate the image color gradation, color difference, and image clarity, and find observers with normal color vision to perform subjective evaluation of memory colors. Through experiments, the evaluation results are closer to the subjective feelings of the human eye, which can better meet customer needs and reduce the cost loss of printed products.
[0026] The highest standard of subjective evaluation is when the printed product is indistinguishable from the original. When there is no original for comparison, the color quality of the printed product is related to the sense of reality, that is, it is required to conform to the color memory of most people. In order to reduce the perceptible loss of image quality, it is necessary to evaluate the color gradation and saturation of the printed image so that the image quality can be maintained, controlled or enhanced during the processing.
[0027] This embodiment provides a method for evaluating printed digital original images based on memory colors, and its process is as follows: Figure 1 As shown, including:
[0028] S1 obtains a printed digital original image.
[0029] The printed digital original image is an RGB image, which is used as the digital image manuscript for various types of printing.
[0030] S2 calculates a first color index.
[0031] Convert the printed digital original image into a LAB image; calculate the average values of L, A and B channels in the LAB image, and calculate the offset between the average value of the L channel and 50, the average value of the A channel and the average value of the B channel to determine the first color index.
[0032] Calculate the average value of each channel to capture the overall color information value of the image, calculate the offset between the average value of the L channel and 50 to get the color deviation b1 to reflect the overall color trend; calculate the Euclidean norm of the average value b2 of the A channel and the average value b3 of the B channel to get the saturation S of the image color a , whose expression is Comprehensively consider the color deviation b1 and saturation S a , and the final first color index X1 is obtained, which is expressed as
[0033] S3 calculates a first clarity index.
[0034] S300 determines the horizontal gradient and the vertical gradient of the printed digital original image, calculates the gradient amplitude value of each pixel based on the horizontal gradient and the vertical gradient, and obtains the maximum amplitude value of the gradient amplitude values as the first maximum amplitude value.
[0035] The gradient is determined by calculating the difference between adjacent pixels in the image, and the maximum gradient represents the maximum rate of change of pixel intensity in the image.
[0036] The horizontal gradient is the rate of change of the image in the horizontal direction, and the vertical gradient is the rate of change of the image in the vertical direction.
[0037] Calculate the horizontal gradient h1 and vertical gradient h2 of the printed digital original image to obtain the gradient amplitude g, whose expression includes
[0038] A first maximum amplitude value MG=max(g) is obtained.
[0039] S310 extracts a central area from the printed digital original image based on the number of rows and columns of pixels contained in the printed digital original image.
[0040] The central area is a square with the same width and height. The expression for determining the size of the central area is:
[0041]
[0042] Among them, B represents a constant to determine the size and position of the central area, m represents the number of rows of pixels contained in the printed digital original image, n represents the number of columns of pixels contained in the printed digital original image, and min(m,n) represents the smaller value between m and n.
[0043] The central area is extracted based on the rows and columns between the B+1th pixel to the Bth pixel from the end in the printed digital original image.
[0044] S320 obtains the mean value, the second maximum value, and the minimum value of the gradient amplitude value in the central area.
[0045] The second maximum amplitude value is the maximum gradient amplitude value in the central area.
[0046] S330 determines the gradient variability value based on the ratio of the difference between the second maximum amplitude value and the minimum amplitude value to the average of the gradient amplitude values in the central area.
[0047] The expression for determining the value of the gradient variability is:
[0048]
[0049] Among them, VG represents the gradient variability value, MaxG represents the second maximum amplitude value, MinG represents the minimum amplitude value, i represents the i-th row of the central area, j represents the j-th column of the central area, G(i,j) represents the pixel value of the i-th row and j-th column of the central area, W represents the width of the central area, and H represents the height of the central area.
[0050] S340 determines a first clarity index based on the first maximum amplitude value and the gradient variability value.
[0051] The first maximum amplitude value and gradient variability are processed using a weighted geometric mean, which is expressed as:
[0052] X2=MG α ×VG 1-α
[0053] Wherein, X2 represents the first clarity index, α represents a parameter, and in this embodiment, α=0.61.
[0054] S4 calculates the first tone index.
[0055] S400 counts the number of occurrences of each pixel value in the printed digital original image, the maximum pixel value and the minimum pixel value in the printed digital original image; and calculates a histogram based on the number of occurrences of each pixel value.
[0056] The probability density function is obtained based on the number of occurrences of each pixel value, and then the cumulative distribution function is calculated to obtain the histogram.
[0057] The horizontal axis of the histogram is the color level, which is the index standard of the brightness of the image, usually corresponding to 0 to 255 from left to right; the vertical axis is the number of pixels at different color levels.
[0058] S410 obtains the first percentile and the third percentile of the histogram; and determines a first tone index based on the ratio of the difference between the third percentile and the first percentile and the difference between the maximum pixel value and the minimum pixel value.
[0059] Percentile: the ratio of the number of pixels within the selected color range on the histogram to the total pixels; the pth percentile in this embodiment means that p% of the data is on the left and (1-p%) is on the right.
[0060] The expressions for determining the tonality index include:
[0061]
[0062] Among them, X3 represents the first tone index, y represents the third percentile of the histogram, x represents the first percentile of the histogram, ma represents the maximum pixel value, and mi represents the minimum pixel value.
[0063] S5: Input the first color index, the first clarity index and the first tone index into the first model to obtain evaluation data of the printed digital original image.
[0064] The expression of the first model is Y=γ1X1′+γ2X2′+γ2X3′, where Y represents the quality score, X1′ represents the color index to be input, X2′ represents the clarity index to be input, X3′ represents the tone index to be input, γ1 represents the first parameter, γ2 represents the second parameter, and γ3 represents the third parameter.
[0065] Determining the first model includes:
[0066] Step 1: Obtain a memory color image data set; randomly sample the memory color image data set, and distort the sampled image to obtain a distorted image, subjectively score the sampled image and the distorted image within a preset time to obtain a scoring score, and calculate a second color index, a second clarity index, and a second tone index.
[0067] The image dataset in the prior art includes reference images and distorted images. The distortion types of the distorted images mainly include white noise distortion, fast Rayleigh attenuation distortion, Gaussian blur, overall contrast reduction distortion, multiple Gaussian noise, etc., and there is no dataset related to color distortion. The purpose of printed materials is to allow people to receive information and make people's eyes more comfortable. The human eye is most sensitive to memory colors. In order to ensure the accuracy and fairness of the research, it is necessary to expand the image dataset. Therefore, this embodiment selects common memory colors to establish an image dataset about memory color distortion for image quality evaluation research.
[0068] The memory color image data set of this embodiment includes 1155 distorted images, of which 273 images are only processed by increasing and decreasing saturation and contrast, and adding 9 kinds of Gaussian noise, and 882 images are randomly distorted, and the distortion processing includes adjusting saturation, contrast or Gaussian noise variance; the score of each image in the memory color image data set is obtained by evaluation, and the larger the score value, the better the quality of the image and the more restored the printing effect. Exemplarily, the display shows the memory color image data set in a dark environment, and 15 observers with normal color vision need to subjectively score the images in the established image data set within 8-15 seconds. The score is the average subjective score of the 15 people, and the score is divided into 1 to 5 points, representing very poor, poor, average, good, and very good, respectively. If it is subjectively considered to be between two levels, it can be scored as a decimal.
[0069] Step 2: Input the scoring score as Y, the second color index as X1′, the second clarity index as X2′, and the second tone index as X3′ into Y=γ1X1′+γ2X3′+γ3X3′ to determine the first parameter, the second parameter, and the third parameter to obtain the first model.
[0070] Preferably, the first parameter is 0.4159, the second parameter is 0.2874, the third parameter is 0.0391, and the expression of the first model is Y=0.4159×X1′+0.28743X2′+0.0391×X3′.
[0071] In summary, the color index, tone index and clarity index are combined, and the three index features are weighted to comprehensively quantify the color loss, contrast reduction and Gaussian noise blur caused by the transmission of images between different devices, which solves the problem of lack of objective evaluation of printed digital manuscripts in the existing image quality evaluation system. Taking full account of the visual perception of the human eye, comprehensively evaluating images from the perspective of color, contrast, clarity and other characteristics, and applying the average opinion score to the image evaluation method, the quality of printed digital manuscripts can be effectively evaluated.
[0072] Compared with the prior art, this embodiment provides a method for evaluating printed digital original images based on memory colors, which objectively reflects whether the printed digital original image is a qualified printed product manuscript, reduces the consumption of manpower and material resources, and its evaluation score is consistent with the visual perception characteristics of the human eye for images. It integrates color, clarity, gradation, and subjective consistency, and combines with the memory color image set to effectively evaluate the color error, noise interference, and gradation change caused by the distortion of image quality during the acquisition, transmission, storage, compression, decompression, and editing of the printed digital original image.
[0073] Example 2
[0074] This embodiment provides a printed digital original image evaluation system based on memory colors, which implements a printed digital original image evaluation method based on memory colors of embodiment 1, including:
[0075] The acquisition module acquires a printed digital original image, wherein the printed digital original image is an RGB image, and serves as a digital image manuscript for various types of printing.
[0076] The color index module calculates the first color index, including: converting the printed digital original image into a LAB image; calculating the average values of the L, A and B channels in the LAB image, and calculating the offset between the average value of the L channel and 50, the average value of the A channel and the average value of the B channel to determine the first color index.
[0077] The clarity index calculation module calculates the first clarity index, including: determining the gradient in the horizontal direction and the gradient in the vertical direction of the printed digital original image, calculating the gradient amplitude value of each pixel based on the gradient in the horizontal direction and the gradient in the vertical direction, and obtaining the maximum amplitude value among the gradient amplitude values as the first maximum amplitude value; extracting the central area from the printed digital original image based on the number of rows and columns of pixels contained in the printed digital original image; obtaining the average value, the second maximum amplitude value and the minimum amplitude value of the gradient amplitude value in the central area, wherein the second maximum amplitude value is the maximum gradient amplitude value in the central area; determining the gradient variability value based on the ratio of the difference between the second maximum amplitude value and the minimum amplitude value to the average value of the gradient amplitude value in the central area; and determining the first clarity index based on the first maximum amplitude value and the gradient variability value.
[0078] The tone index calculation module calculates the first tone index, including: counting the number of occurrences of each pixel value in the printed digital original image, the maximum pixel value and the minimum pixel value in the printed digital original image, and calculating a histogram based on the number of occurrences of each pixel value; obtaining the first percentile and the third percentile of the histogram; and determining the first tone index based on the difference between the third percentile and the first percentile and the ratio of the difference between the maximum pixel value and the minimum pixel value.
[0079] The output module inputs the first color index, the first clarity index and the first tone index into the first model to obtain evaluation data of the printed digital original image. The expression of the first model is Y=γ1X1′+γ2X2′+γ3X3′, wherein Y represents the quality score, X1′ represents the color index to be input, X2′ represents the clarity index to be input, X3′ represents the tone index to be input, γ1 represents the first parameter, γ2 represents the second parameter, and γ3 represents the third parameter.
[0080] Optionally, determining the first model includes:
[0081] Acquire a memory color image data set; randomly sample the memory color image data set, and perform distortion processing on the sampled image to obtain a distorted image, wherein the distortion processing includes adjusting saturation, contrast or Gaussian noise variance; within a preset time, subjectively score the sampled image and the distorted image to obtain a score, and calculate a second color index, a second clarity index and a second tone index; input the score as Y, the second color index as X1′, the second clarity index as X2′, and the second tone index as X3′ into Y=γ1X1′+γ2X2′+γ3X3′ to determine the first parameter, the second parameter and the third parameter, and obtain a first model.
[0082] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for evaluating printed digital original images based on memory colors, characterized in that: include: Acquire a printed digital original image, wherein the printed digital original image is an RGB image, which is used as a digital image manuscript for various types of printing; Calculating a first color index includes: converting a printed digital original image into a LAB image; calculating average values of L, A, and B channels in the LAB image, and calculating an offset between the average value of the L channel and 50, an average value of the A channel, and an average value of the B channel to determine the first color index; Calculating a first definition index includes: determining a gradient in a horizontal direction and a gradient in a vertical direction of a printed digital original image, calculating a gradient amplitude value of each pixel based on the gradient in the horizontal direction and the gradient in the vertical direction, and obtaining a maximum amplitude value among the gradient amplitude values as a first maximum amplitude value; extracting a central area from the printed digital original image based on the number of rows and columns of pixels contained in the printed digital original image; obtaining an average value, a second maximum amplitude value, and a minimum amplitude value of the gradient amplitude values in the central area, wherein the second maximum amplitude value is the maximum gradient amplitude value in the central area; determining a gradient variability value based on a ratio of a difference between the second maximum amplitude value and the minimum amplitude value to a mean value of the gradient amplitude values in the central area; and determining a first definition index based on the first maximum amplitude value and the gradient variability value; Calculating the first tone index includes: counting the number of occurrences of each pixel value in the printed digital original image, the maximum pixel value and the minimum pixel value in the printed digital original image, and calculating a histogram based on the number of occurrences of each pixel value; obtaining a first percentile and a third percentile of the histogram; determining the first tone index based on a difference between the third percentile and the first percentile and a ratio of a difference between the maximum pixel value and the minimum pixel value; The first color index, the first clarity index and the first tone index are input into the first model to obtain evaluation data of the printed digital original image. The expression of the first model is Y=γ1X1′+γ2X3′+γ3X3′, wherein Y represents the quality score, X1′ represents the color index to be input, X2′ represents the clarity index to be input, X3′ represents the tone index to be input, γ1 represents the first parameter, γ2 represents the second parameter, and γ3 represents the third parameter.
2. The method for evaluating printed digital original images based on memory colors according to claim 1, characterized in that: The offset between the average value of the L channel and 50, the average value of the A channel and the average value of the B channel are calculated to determine the first color index, whose expression includes: Among them, S a Indicates the saturation of the color, b1 indicates the offset between the average value of channel L and 50, b2 indicates the average value of channel A, b3 indicates the average value of channel B, and X1 indicates the first color index.
3. The method for evaluating printed digital original images based on memory colors according to claim 1, characterized in that: Extracting a central area from the printed digital original image based on the number of rows and columns of pixels contained in the printed digital original image includes: Wherein, m represents the number of rows of pixels contained in the printed digital original image, n represents the number of columns of pixels contained in the printed digital original image, and B represents a constant to determine the size and position of the central area; The central area is extracted based on the rows and columns between the B+1th pixel to the Bth pixel from the end in the printed digital original image.
4. The method for evaluating printed digital original images based on memory colors according to claim 1, characterized in that: Based on the ratio of the difference between the second maximum amplitude value and the minimum amplitude value to the average value of the gradient amplitude value in the central area, the gradient variability value is determined, and its expression includes: Among them, VG represents the gradient variability value, MaxG represents the second maximum amplitude value, MinG represents the minimum amplitude value, i represents the i-th row of the central area, j represents the j-th column of the central area, G(i,j) represents the value of the i-th row and j-th column of the central area, W represents the width of the central area, and H represents the height of the central area.
5. The method for evaluating printed digital original images based on memory colors according to claim 1, characterized in that: A first clarity index is determined based on the first maximum amplitude value and the gradient variability value, and the expression thereof includes: X2=MG α ×VG 1-α Among them, X2 represents the first clarity index, MG represents the first maximum amplitude value, VG represents the gradient variability value, and α represents a constant.
6. The method for evaluating printed digital original images based on memory colors according to claim 1, characterized in that: The expression for calculating the first tone index includes: Among them, X3 represents the first tone index, y represents the third percentile of the histogram, x represents the first percentile of the histogram, ma represents the maximum pixel value, and mi represents the minimum pixel value.
7. The method for evaluating printed digital original images based on memory colors according to claim 1, characterized in that: Determining the first model includes: Acquire a memory color image data set; randomly sample the memory color image data set, and perform distortion processing on the sampled image to obtain a distorted image, wherein the distortion processing includes adjusting saturation, contrast or Gaussian noise variance; within a preset time, subjectively score the sampled image and the distorted image to obtain a score, and calculate a second color index, a second clarity index and a second tone index; input the score as Y, the second color index as X1′, the second clarity index as X2′, and the second tone index as X3′ into Y=γ1X1′+γ2X2′+γ3X3′ to determine the first parameter, the second parameter and the third parameter, and obtain a first model.
8. The method for evaluating printed digital original images based on memory colors according to claim 7, characterized in that: The expression of the first model is: Y=0.4159×X1′+0.2874×X2′+0.0391×X3′.
9. A printed digital original image evaluation system based on memory color, characterized in that: include: An acquisition module is used to acquire a printed digital original image, wherein the printed digital original image is an RGB image, and serves as a digital image manuscript for various types of printing; The color index module calculates the first color index, including: converting the printed digital original image into a LAB image; calculating the average values of L, A and B channels in the LAB image, and calculating the offset between the average value of the L channel and 50, the average value of the A channel and the average value of the B channel to determine the first color index; The clarity index calculation module calculates the first clarity index, including: determining the gradient in the horizontal direction and the gradient in the vertical direction of the printed digital original image, calculating the gradient amplitude value of each pixel based on the gradient in the horizontal direction and the gradient in the vertical direction, and obtaining the maximum amplitude value among the gradient amplitude values as the first maximum amplitude value; extracting the central area from the printed digital original image based on the number of rows and columns of pixels contained in the printed digital original image; obtaining the average value, the second maximum amplitude value and the minimum amplitude value of the gradient amplitude value of the central area, wherein the second maximum amplitude value is the maximum gradient amplitude value in the central area; determining the gradient variability value based on the ratio of the difference between the second maximum amplitude value and the minimum amplitude value to the average value of the gradient amplitude value of the central area; and determining the first clarity index based on the first maximum amplitude value and the gradient variability value; The tone index calculation module calculates the first tone index, including: counting the number of occurrences of each pixel value in the printed digital original image, the maximum pixel value and the minimum pixel value in the printed digital original image, and calculating a histogram based on the number of occurrences of each pixel value; obtaining the first percentile and the third percentile of the histogram; and determining the first tone index based on the ratio of the difference between the third percentile and the first percentile and the difference between the maximum pixel value and the minimum pixel value; The output module inputs the first color index, the first clarity index and the first tone index into the first model to obtain evaluation data of the printed digital original image. The expression of the first model is Y=γ1X1′+γ2X2′+γ3X3′, wherein Y represents the quality score, X1′ represents the color index to be input, X2′ represents the clarity index to be input, X3′ represents the tone index to be input, γ1 represents the first parameter, γ2 represents the second parameter, and γ3 represents the third parameter.
10. The printed digital original image evaluation system based on memory color according to claim 9, characterized in that: Determining the first model includes: Acquire a memory color image data set; randomly sample the memory color image data set, and perform distortion processing on the sampled image to obtain a distorted image, wherein the distortion processing includes adjusting saturation, contrast or Gaussian noise variance; within a preset time, subjectively score the sampled image and the distorted image to obtain a score, and calculate a second color index, a second clarity index and a second tone index; input the score as Y, the second color index as X1′, the second clarity index as X2′, and the second tone index as X3′ into Y=γ1X1′+γ2X2′+γ3X3′ to determine the first parameter, the second parameter and the third parameter, and obtain a first model.
