Human eye color vision defect quantitative detection method and detection device

Abstract

The invention relates to human eye color vision detection, in particular to a human eye color vision defect quantitative detection method and a human eye color vision defect quantitative detection device, which are characterized in that a color resolution limit of a subject is gradually approached by using a time and space two-dimensional alternate change detection pattern in combination with a dichotomy chromaticity threshold interactive feedback test method, and a Mecanum color confusion ellipse of the subject is obtained; inputting the Mecanum color confusion ellipse of the subject into the trained color vision defect type identification model to obtain the color vision defect type of the subject; determining a changed LMS cone spectral response curve according to the color vision defect type of the subject, and obtaining an LMS cone spectral response curve of the subject in combination with the LMS cone spectral response curve of the standard observer; according to the technical scheme provided by the invention, the defect that the human eye color vision defect is difficult to accurately and quantitatively detect in the prior art can be effectively overcome.

Description

Technical Field

[0001] The invention relates to human eye color vision detection, and in particular to a method and a device for quantitatively detecting human eye color vision defects. Background Art

[0002] At present, the existing methods for detecting color vision defects of the human eye mainly include pseudo-isochromatic diagrams, hue arrangement methods, color vision mirrors and genetic testing.

[0003] Pseudoisochromatic patterns are patterns with spots of equal brightness but easily confused colors forming the background and foreground, and contain information such as text and graphics. People with color vision abnormalities can only judge by brightness and darkness, and cannot distinguish the color difference between the background and foreground, and thus cannot distinguish the information in the pattern. This detection method has three main problems: 1) This method can only perform qualitative detection, and it is difficult to distinguish the type and severity of the subject's color vision defect. It has been controversial in the decades since its birth; 2) This method is greatly affected by the light source, and is generally required to be carried out under the lighting conditions of the D65 standard light source, but it is difficult to ensure the consistency of the use environment during actual use; 3) After training, the subjects can remember the true information of the test pattern based on the geometric shape arrangement, brightness changes and other information of the test pattern to resist inspection.

[0004] The hue arrangement method uses a series of color particles with continuous hues. Each hue particle has a number on the back. After the hue particles are shuffled into a disordered state, the subjects are asked to rearrange them according to their color vision discrimination ability. The examiner compares the arranged number order with that of normal people to determine whether the subjects have color vision abnormalities and the type of color vision abnormalities. This detection method has three main problems: 1) For situations where detailed detection results are required, the inspection time is long and it is difficult to conduct a large-scale census; 2) Some studies have shown that for people with mild color vision defects, the accuracy of their hue arrangement is not much different from that of normal people, and the accuracy can be improved after training, and there is also the situation of confrontational inspection; 3) Like the false isochromatic diagram, this method also needs to be carried out under the lighting conditions of the D65 standard light source.

[0005] The color vision mirror is designed based on the principle of color matching. It is currently the most accurate color vision quantitative and qualitative test solution and is known as the "gold standard" in the field of color vision defect detection. However, this detection method has problems such as expensive equipment, long inspection time, and long inspection time makes the inspector and the subject very easy to become fatigued, which affects the accuracy of the test results.

[0006] Genetic testing determines whether a subject has color vision defects by detecting the gene sequence related to color. This testing method has a high accuracy in qualitative examination of congenital color vision defects, but the testing cost is high and it cannot give quantitative test results. At the same time, it cannot give test results for acquired color vision defects, such as color vision defects caused by lens degeneration caused by diseases such as cataracts and diabetes.

[0007] The invention patent application for a refined color vision detection system based on multi-color pair matching, with application publication number CN117717308A, is based on the principle of color vision mirror and has developed a system that can infer the LMS (L for long-wave sensitivity, M for medium-wave sensitivity, and S for short-wave sensitivity) cone spectral response curve of the human eye. However, the system also has the problem of long inspection time, and relies on the peak wavelength shift of the L curve, the peak wavelength shift of the M curve, and the peak intensity change of the S curve to determine the type of color vision abnormality of the subject. The detection results are very dependent on the pre-set thresholds of each parameter, and the reliability of the detection results has not been effectively verified. Summary of the invention

[0008] 1. Technical issues to be resolved

[0009] In view of the above-mentioned shortcomings of the prior art, the present invention provides a method and a device for quantitatively detecting color vision defects of the human eye, which can effectively overcome the shortcomings of the prior art that it is difficult to accurately and quantitatively detect color vision defects of the human eye.

[0010] (II) Technical solution

[0011] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0012] A method for quantitatively detecting color vision defects of human eyes comprises the following steps:

[0013] S1. Using the detection pattern that changes alternately in two dimensions of time and space, combined with the dichotomy chromaticity threshold interactive feedback test method, gradually approaching the color discrimination limit of the subject, and obtaining the subject's MacAdam color confusion ellipse;

[0014] S2, inputting the subject's MacAdam color confusion ellipse into the trained color vision deficiency type recognition model to obtain the subject's color vision deficiency type;

[0015] S3. Determine the changed LMS cone spectral response curve according to the type of color vision deficiency of the subject, and obtain the LMS cone spectral response curve of the subject by combining the LMS cone spectral response curve of the standard observer.

[0016] Preferably, in S1, a detection pattern that alternately changes in two dimensions of time and space is used, combined with a binary chromaticity threshold interactive feedback test method, to gradually approach the color discrimination limit of the subject, and obtain the subject's MacAdam color confusion ellipse, including:

[0017] The angles determined by the red, green, and blue confusion lines of the background color chromaticity coordinates of the detection pattern are used as sampling angles to obtain the six points that best reflect the color vision problem to approximately fit the subject's MacAdam color confusion ellipse;

[0018] The number of steps n corresponding to the starting minimum boundary of the binary division in each sampling direction min is 0.0001 chromaticity coordinate unit, that is, n min =1, the number of steps corresponding to the starting highest boundary n max is the Euclidean distance between the sRGB boundary and the background coordinate point in this direction, and the step size Δl is 0.0001 chromaticity coordinate unit;

[0019] The number of steps n corresponding to the lowest boundary is continuously updated during the process of the subject's feedback on the direction of the dynamic sight mark movement. min The number of steps n corresponding to the highest boundary max , when |n max -n min |≤1, and the subject cannot identify the direction of motion of the dynamic sight mark for multiple consecutive times, the direction test is terminated, and the step length corresponding to the highest boundary obtained is n max It is defined as the chromaticity discrimination threshold of the subject at the sampling angle;

[0020] Among them, the red, green and blue confusion lines are defined as straight lines determined by the intersection points of all red, green and blue confusion lines and the background color chromaticity coordinates. In the CIE 1976L*u*v* chromaticity coordinate system, the intersection point of all red confusion lines is (0.678, 0.501), the intersection point of all green confusion lines is (-1.217, 0.782), and the intersection point of all blue confusion lines is (0.257, 0). When the background point is the chromaticity corresponding to the D65 standard light source, that is, the background color chromaticity coordinates are (0.1978, 0.4684), the angles of the red, green and blue confusion lines relative to the positive direction of the u' coordinate axis are 3.8838° or 183.8838°, -12.4979° or 167.5021°, and -82.7967° or 97.2033°, respectively, and there are 6 intersection points with the MacAdam color confusion ellipse.

[0021] Preferably, the detection pattern uses a moving dynamic sight mark instead of a static sight mark, converting color vision detection into directional feedback, thereby avoiding the loophole of the conventional detection method in which the subject memorizes the detection pattern after training from the detection mechanism;

[0022] Random brightness noise is added to the background color and the sight mark color of the detection pattern to avoid the loophole that the subject uses the brightness information in the detection pattern to identify the sight mark, thereby improving the reliability of the test result;

[0023] The dynamic sight mark moves from the lower right corner to the upper left corner of the background image, or from the lower left corner to the upper right corner. The subject provides feedback on the movement direction of the dynamic sight mark through an input device during or after the movement of the dynamic sight mark.

[0024] Preferably, the background color chromaticity coordinates (u 0 ,v 0 ) is set as any point in the CIE 1976 L*u*v* chromaticity coordinate system, and the average brightness value of all small squares in the background image is 24cd / m 2 , the brightness noise is ±45%, and the background image covers the field of view of the human eye of more than 10°*10°;

[0025] The visual standard chromaticity coordinates (u t ,v t ) is calculated using the following formula:

[0026]

[0027] Where ΔL is the Euclidean distance of the current test, and θ is the angle of the current test;

[0028] All the small squares in the dynamic sight mark correspond to the same brightness value, the brightness mean and brightness noise of all the small squares in the dynamic sight mark are the same as the background image, and the dynamic sight mark covers a 2°*2° field of view of the human eye.

[0029] Preferably, before inputting the subject's MacAdam color confusion ellipse into a trained color vision deficiency type recognition model to obtain the subject's color vision deficiency type in S2, the method includes:

[0030] S21, collecting MacAdam color confusion ellipses and corresponding color vision deficiency types of a large number of subjects, annotating the MacAdam color confusion ellipses using the color vision deficiency types, constructing a historical data set, and dividing the historical data set into a training set, a validation set, and a test set according to a preset ratio;

[0031] S22, setting a loss function and an optimizer for the color vision deficiency type recognition model;

[0032] S23, inputting the training set into the color vision deficiency type recognition model for model training;

[0033] S24, calculating the loss value based on the loss function, and the optimizer updates the model parameters according to the loss value and the network gradient information;

[0034] S25. If the loss value is less than the preset threshold, the model training ends, and the current color vision deficiency type recognition model is the trained color vision deficiency type recognition model. Otherwise, the process returns to S23 and continues to use the training set for model training.

[0035] S26. Input the validation set into the trained color vision deficiency type recognition model, evaluate the generalization ability of the model by observing the performance on the validation set, and tune the hyperparameters and structure of the model;

[0036] S27. Input the test set into the optimized color vision deficiency type recognition model to evaluate the model performance.

[0037] Preferably, determining the changed LMS cone spectral response curve according to the color vision defect type of the subject in S3, and obtaining the LMS cone spectral response curve of the subject in combination with the LMS cone spectral response curve of a standard observer, comprises:

[0038] S31, test to obtain the subject's MacAdam color confusion threshold chromaticity coordinates (u ti ,v ti ), i∈[1,6];

[0039] S32, test or retrieve the background color chromaticity coordinates (u 0 ,v 0 ), MacAdam color confusion threshold chromaticity coordinates (u ti ,v ti ) corresponds to the spectral SPD 0 (λ), SPD ti (λ);

[0040] S33, using the LMS cone spectral response curve lms of the standard observer standard (λ) Calculate the color confusion tristimulus value (X) of the standard observer in the CIE 1931XYZ tristimulus value coordinate system ti ,Y ti ,Z ti ), background point tristimulus values ​​of the standard observer (X 0 ,Y 0 ,Z 0 );

[0041] S34. Calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ) and the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 ;

[0042] S35. Determine the LMS cone spectral response curve lms according to the type of color vision deficiency of the subject calculate (λ), using the varying LMS cone spectral response curve lms calculate (λ) Calculate the color confusion tristimulus value (X') in the CIE 1931XYZ tristimulus value coordinate system ti ,Y' ti ,Z' ti ), the changing background point tristimulus value (X' 0 ,Y' 0 ,Z' 0 );

[0043] S36, calculate the color confusion tristimulus value (X' ti ,Y' ti ,Z' ti ) and the changing background point tristimulus value (X' 0 ,Y' 0 ,Z' 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 ';

[0044] S37, if ΔE 00 '<ε, then the LMS cone spectral response curve lms is continuously optimized. calculate (λ) parameter, so that |ΔE 00 '-ΔE 00 |<ζ, if ΔE 00 '≥ε or |ΔE 00 '-ΔE 00 |≥ζ, then return to S35;

[0045] S38, when |ΔE 00 '-ΔE 00 When |<ζ, the changing LMS cone spectral response curve lms can be obtained. calculate The parameters of (λ) are used to obtain the LMS cone spectral response curve of the subject;

[0046] Among them, the LMS cone spectral response curve of the standard observer is lms standard (λ) is the cone spectral response curve of the standard observer specified in CIE 170-1-2006 obtained when the subject age a=32, the field angle v=2°, and other parameters are all 0 in the LMS cone spectral response mathematical model;

[0047] Variation of the LMS cone spectral response curve lms calculate(λ) is obtained by taking random values ​​of parameters in the mathematical model of LMS cone spectral response, and the range of values ​​of each parameter is determined according to the type of color vision deficiency of the subject;

[0048] ε is the color difference threshold, ζ is the color difference difference threshold, and ε and ζ are determined according to the type of color vision deficiency of the subject.

[0049] Preferably, the LMS cone spectral response mathematical model lms(λ,j) is expressed by the following formula:

[0050] lms(λ,j)=f(λ,j,a,v,d lens ,d macula ,d L ,d M ,d S ,s L ,s M ,s S );

[0051] Where λ is the wavelength, j is the cone cell type, j = L, M, S, representing L cone cells, M cone cells, and S cone cells, respectively, a is the age of the subject, v is the field of view, v = 2°, d lens is the deviation of the peak value of the optical density spectrum of the lens and other intraocular media relative to that of a standard observer, d macula is the deviation of the mean value of the macular pigment optical density spectrum relative to the standard observer;

[0052] d L ,d M ,d S are the deviations of the peak values ​​of the optical density spectra of the L, M, and S cone cells relative to the standard observer, s L 、s M 、s S are the deviations of the peak wavelengths of the optical density spectra of the L, M, and S cone pigments relative to the standard observer;

[0053] The normalized LMS cone spectral response function in quanta is:

[0054]

[0055] in, are the normalized LMS cone spectral response functions of L, M, and S cone cells in quantum units, α l (λ), α m (λ), α s (λ) are the absorption spectra of the pigments of L, M, and S cone cells, respectively, and D macula (λ) is the optical density spectrum of macular pigment, D ocul (λ) is the optical density spectrum of the lens and other intraocular media;

[0056] The normalized LMS cone spectral response function in energy units is:

[0057]

[0058] in, are the normalized LMS cone spectral response functions of L, M, and S cone cells in energy units, respectively, and max[·] represents the calculated peak value.

[0059] Preferably, the color confusion tristimulus values ​​(X ti ,Y ti ,Z ti ), the background point tristimulus value of the standard observer (X 0 ,Y 0 ,Z 0 ), changing color confusion tristimulus value (X' ti ,Y' ti ,Z' ti ) and the changing background point tristimulus value (X' 0 ,Y' 0 ,Z' 0 ) are collectively referred to as tristimulus values ​​(X, Y, Z), and the tristimulus values ​​(X, Y, Z) are calculated using the following formula:

[0060]

[0061] Where SPD(λ) is the chromaticity coordinate of the background color (u 0 ,v 0 ) corresponds to the spectral SPD 0 (λ), or the MacAdam color confusion threshold chromaticity coordinate (u ti ,v ti ) corresponds to the spectral SPD ti (λ).

[0062] Preferably, in S34, the color confusion tristimulus values ​​(X ti ,Y ti ,Z ti ) and the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 ,include:

[0063] S341. Calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ), the background point tristimulus value of the standard observer (X 0 ,Y0 ,Z 0 ) in the CIE 1976 L*u*v* chromaticity coordinate system L*, a*, b* and

[0064]

[0065] Among them, (X, Y, Z) represents the color confusion tristimulus value of the standard observer (X ti ,Y ti ,Z ti ) or the background point tristimulus value of the standard observer (X 0 ,Y 0 ,Z 0 ), (X n ,Y n ,Z n ) represents the tristimulus value of D65 light source, X n =95.047, Y n =100.000, Z n =108.883;

[0066] S342. Calculate the following intermediate variables:

[0067]

[0068] in, is the arithmetic mean of the chroma values ​​of the background point and the threshold point;

[0069] S343. Calculate the following intermediate variables:

[0070]

[0071] in, is the color confusion tristimulus value of the standard observer (X ti ,Y ti ,Z ti )L*, C′ in the CIE 1976 L*u*v* chromaticity coordinate system 1 is the color confusion tristimulus value (X ti ,Y ti ,Z ti ) Calculated C', h' 1 is the color confusion tristimulus value (X ti ,Y ti ,Z ti ) calculated h';

[0072] is the background point tristimulus value of the standard observer (X 0 ,Y0 ,Z 0 )L*, C′ in the CIE 1976 L*u*v* chromaticity coordinate system 2 is the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Calculated C', h' 2 is the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) calculated h';

[0073] S344, calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ) and the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 :

[0074]

[0075] in,

[0076]

[0077] R T = -sin(2·Δθ)·R C

[0078]

[0079] for and The average value of C′ 1 and C′ 2 The average value of h′ 1 and h′ 2 The average value of .

[0080] A human eye color vision defect detection device, comprising a computer, a colorimeter, a spectrometer, a display and an input device;

[0081] A colorimeter is used to correct the color difference of the display to within the allowable range. When correcting the color difference, the colorimeter generates a source color calibration pattern and sends it to the display, and at the same time receives the color calibration pattern displayed by the display, and compares the color calibration pattern with the source color calibration pattern, calculates the color calibration file, and sends it to the computer;

[0082] A spectrometer is used to collect spectral information of the test pattern. When the chromaticity of the display pattern is the same as the chromaticity of the D65 light source, the display brightness is calculated based on the spectral information collected by the spectrometer to adjust the display brightness during the test;

[0083] A display, used to display a calibration pattern to the colorimeter, transmit spectral information of the detection pattern to the spectrometer, display the detection pattern and the detection result to the subject, and display operation interaction information;

[0084] The computer controls the operation of the colorimeter, the spectrometer and the display, and generates the detection pattern and sends it to the display for display;

[0085] An input device used for the subject to provide feedback on the direction of movement of the dynamic sight mark during or after the movement of the dynamic sight mark.

[0086] (III) Beneficial effects

[0087] Compared with the prior art, the present invention provides a quantitative detection method and device for color vision defects of human eyes, which utilize a detection pattern that changes alternately in two dimensions of time and space, combined with a binary chromaticity threshold interactive feedback test method, to gradually approach the color discrimination limit of the subject, obtain the subject's MacAdam color confusion ellipse, and obtain the subject's color vision defect type through a color vision defect type recognition model, determine the changed LMS cone spectral response curve according to the subject's color vision defect type, and obtain the subject's LMS cone spectral response curve in combination with the LMS cone spectral response curve of a standard observer, thereby enabling accurate quantitative detection of the subject's human eye color vision defect. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0089] Figure 1 is a schematic diagram of a detection pattern in the present invention;

[0090] Figure 2 It is a flow chart of the binary chromaticity threshold interactive feedback testing method of the present invention;

[0091] Figure 3 is the MacAdam color confusion ellipse of a normal subject in the present invention;

[0092] Figure 4 is the MacAdam color confusion ellipse of the subject with red color vision deficiency in the present invention;

[0093] Figure 5 is the MacAdam color confusion ellipse of the green color vision deficient subject in the present invention;

[0094] Figure 6 is the MacAdam color confusion ellipse for the subject with blue color vision deficiency in the present invention;

[0095] Figure 7 It is a schematic diagram of a process of obtaining the LMS cone spectral response curve of a subject according to the color vision deficiency type of the subject in the present invention;

[0096] Figure 8 Schematic diagram of the hardware composition of the human eye color vision defect detection device in the present invention;

[0097] Fig. 9 To adopt Figure 7 The method shown obtains the LMS cone spectral response curve of normal subjects;

[0098] Fig.10 To adopt Figure 7 The method shown obtains the LMS cone spectral response curve of the red color vision deficient subject;

[0099] Fig.11 To adopt Figure 7 The method shown obtains the LMS cone spectral response curve of the green color vision deficient subject;

[0100] Fig.12 To adopt Figure 7 The method shown obtains the LMS cone spectral response curve of a subject with blue color vision deficiency. DETAILED DESCRIPTION

[0101] 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.

[0102] A quantitative detection method for color vision deficiency of human eyes, S1, using a detection pattern that changes alternately in two dimensions of time and space, combined with a dichotomy chromaticity threshold interactive feedback test method, gradually approaching the color discrimination limit of a subject, and obtaining the subject's MacAdam color confusion ellipse, specifically comprising:

[0103] The angles determined by the red, green, and blue confusion lines of the background color chromaticity coordinates of the detection pattern are used as sampling angles to obtain the six points that best reflect the color vision problem to approximately fit the subject's MacAdam color confusion ellipse;

[0104] The number of steps n corresponding to the starting minimum boundary of the binary division in each sampling direction min is 0.0001 chromaticity coordinate unit, that is, n min =1, the number of steps corresponding to the starting highest boundary n max is the Euclidean distance between the sRGB boundary and the background coordinate point in this direction, that is, n max =n B , the step length Δl is 0.0001 chromaticity coordinate unit;

[0105] The number of steps n corresponding to the lowest boundary is continuously updated during the process of the subject's feedback on the direction of the dynamic sight mark movement. min The number of steps n corresponding to the highest boundary max , when |n max -n min |≤1, and the subject cannot identify the direction of motion of the dynamic sight mark for multiple consecutive times, the direction test is terminated, and the step length corresponding to the highest boundary obtained is n max It is defined as the chromaticity discrimination threshold of the subject at the sampling angle;

[0106] Among them, the red, green and blue confusion lines are defined as the straight lines determined by the intersection points of all red, green and blue confusion lines and the background color chromaticity coordinates. In the CIE 1976L*u*v* chromaticity coordinate system, the intersection point of all red confusion lines is (0.678, 0.501), the intersection point of all green confusion lines is (-1.217, 0.782), and the intersection point of all blue confusion lines is (0.257, 0). When the background point is the chromaticity corresponding to the D65 standard light source, that is, the background color chromaticity coordinate is (0.1978, 0.4684), the angles of the red, green and blue confusion lines relative to the positive direction of the u' coordinate axis are 3.8838° or 183.8838°, -12.4979° or 167.5021°, and -82.7967° or 97.2033°, respectively. There are 6 intersection points with the MacAdam color confusion ellipse, such as Figures 3 to 6 shown.

[0107] like Figure 1 As shown, the detection pattern uses a moving dynamic sight mark instead of a static sight mark, converting color vision detection into directional feedback, and avoiding the loophole of the conventional detection method in which the subjects memorize the detection pattern after training from the detection mechanism;

[0108] Random brightness noise is added to the background color and visual mark color of the test pattern to avoid the loophole that the subject uses the brightness information in the test pattern to identify the visual mark, thus improving the reliability of the test results;

[0109] The dynamic sight mark moves from the lower right corner to the upper left corner of the background image, or from the lower left corner to the upper right corner. The subject provides feedback on the movement direction of the dynamic sight mark through an input device during or after the movement of the dynamic sight mark.

[0110] like Figure 1 As shown, the background color chromaticity coordinates (u 0 ,v 0 ) is set to any point in the CIE 1976L*u*v* chromaticity coordinate system, and the average brightness value of all small squares in the background image is 24cd / m 2 , the brightness noise is ±45%, and the background image covers the field of view of the human eye of more than 10°*10°;

[0111] The visual standard chromaticity coordinates of the detection pattern (u t ,v t ) is calculated using the following formula:

[0112]

[0113] Where ΔL is the Euclidean distance of the current test, and θ is the angle of the current test;

[0114] All small squares in the dynamic sight mark correspond to the same brightness value. The brightness mean and brightness noise of all small squares in the dynamic sight mark are the same as the background image. The dynamic sight mark covers the 2°*2° field of view of the human eye.

[0115] Figure 2 Schematic diagram of the flow of the binary color threshold interactive feedback test method. temp Indicates the number of steps of the current test, n min Indicates the number of steps corresponding to the lowest boundary of the next test, n max Indicates the number of steps corresponding to the highest boundary of the next test, n B Indicates the number of chromaticity coordinate steps corresponding to the intersection of the sRGB boundary and the color confusion line currently being tested, Δl represents the step length, θ represents the angle currently being tested, and n error Indicates the direction feedback error count.

[0116] Figures 3 to 6 The MacAdam color confusion ellipses are for normal subjects, subjects with red color vision deficiency, subjects with green color vision deficiency, and subjects with blue color vision deficiency. 1 is the red confusion line, 2 is the green confusion line, 3 is the blue confusion line, 4 is the boundary of the CIE 1976L*u*v* chromaticity coordinate system, 5 is the sRGB displayable chromaticity boundary used by mainstream monitors, 6 is the measured chromaticity coordinate threshold of the subject, 7 is the fitted boundary of the MacAdam color confusion ellipse of the subject, and 8 is the upper limit of the MacAdam color confusion boundary of the normal subject (exceeding this boundary means that the color vision is defective).

[0117] S2. Input the subject's MacAdam color confusion ellipse into the trained color vision deficiency type recognition model to obtain the subject's color vision deficiency type.

[0118] Specifically, before inputting the subject's MacAdam color confusion ellipse into the trained color vision deficiency type recognition model to obtain the subject's color vision deficiency type, the following steps are included:

[0119] S21, collecting MacAdam color confusion ellipses and corresponding color vision deficiency types of a large number of subjects, annotating the MacAdam color confusion ellipses using the color vision deficiency types, constructing a historical data set, and dividing the historical data set into a training set, a validation set, and a test set according to a preset ratio;

[0120] S22, setting a loss function and an optimizer for the color vision deficiency type recognition model;

[0121] S23, inputting the training set into the color vision deficiency type recognition model for model training;

[0122] S24, calculating the loss value based on the loss function, and the optimizer updates the model parameters according to the loss value and the network gradient information;

[0123] S25. If the loss value is less than the preset threshold, the model training ends, and the current color vision deficiency type recognition model is the trained color vision deficiency type recognition model. Otherwise, the process returns to S23 and continues to use the training set for model training.

[0124] S26. Input the validation set into the trained color vision deficiency type recognition model, evaluate the generalization ability of the model by observing the performance on the validation set, and tune the hyperparameters and structure of the model;

[0125] S27. Input the test set into the optimized color vision deficiency type recognition model to evaluate the model performance.

[0126] S3. Determine the changed LMS cone spectral response curve according to the type of color vision deficiency of the subject, and obtain the LMS cone spectral response curve of the subject by combining the LMS cone spectral response curve of the standard observer, such as Figure 7 As shown, specifically including:

[0127] S31, test to obtain the subject's MacAdam color confusion threshold chromaticity coordinates (u ti ,v ti ), i∈[1,6];

[0128] S32, test or retrieve the background color chromaticity coordinates (u 0 ,v 0 ), MacAdam color confusion threshold chromaticity coordinates (u ti ,v ti) corresponds to the spectral SPD 0 (λ), SPD ti (λ);

[0129] S33, using the LMS cone spectral response curve lms of the standard observer standard (λ) Calculate the color confusion tristimulus value (X) of the standard observer in the CIE 1931XYZ tristimulus value coordinate system ti ,Y ti ,Z ti ), the background point tristimulus value of the standard observer (X 0 ,Y 0 ,Z 0 );

[0130] S34. Calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ) and the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 ;

[0131] S35. Determine the LMS cone spectral response curve lms according to the type of color vision deficiency of the subject calculate (λ), using the varying LMS cone spectral response curve lms calculate (λ) Calculate the color confusion tristimulus value (X') in the CIE 1931XYZ tristimulus value coordinate system ti ,Y' ti ,Z' ti ), the changing background point tristimulus value (X' 0 ,Y' 0 ,Z' 0 );

[0132] S36, calculate the color confusion tristimulus value (X' ti ,Y' ti ,Z' ti ) and the changing background point tristimulus value (X' 0 ,Y' 0 ,Z' 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 ';

[0133] S37, if ΔE 00 '<ε, then the LMS cone spectral response curve lms is continuously optimized. calculate (λ) parameter, so that |ΔE00 '-ΔE 00 |<ζ, if ΔE 00 '≥ε or |ΔE 00 '-ΔE 00 |≥ζ, then return to S35;

[0134] S38, when |ΔE 00 '-ΔE 00 When |<ζ, the changing LMS cone spectral response curve lms can be obtained. calculate The parameters of (λ) are used to obtain the LMS cone spectral response curve of the subject, such as Figures 9 to 12 As shown;

[0135] Among them, the LMS cone spectral response curve of the standard observer is lms standard (λ) is the cone spectral response curve of the standard observer specified in CIE 170-1-2006 obtained when the subject age a=32, the field angle v=2°, and other parameters are all 0 in the LMS cone spectral response mathematical model;

[0136] Variation of the LMS cone spectral response curve lms calculate (λ) is obtained by taking random values ​​of parameters in the mathematical model of LMS cone spectral response, and the range of values ​​of each parameter is determined according to the type of color vision deficiency of the subject;

[0137] ε is the color difference threshold, ζ is the color difference difference threshold, and ε and ζ are determined according to the type of color vision deficiency of the subject.

[0138] The mathematical model of LMS cone spectral response lms(λ,j) is expressed as follows:

[0139] lms(λ,j)=f(λ,j,a,v,d lens ,d macula ,d L ,d M ,d S ,s L ,s M ,s S );

[0140] Where λ is the wavelength, j is the cone cell type, j = L, M, S, representing L cone cells, M cone cells, and S cone cells, respectively, a is the age of the subject, v is the field of view, v = 2°, d lens is the deviation of the peak value of the optical density spectrum of the lens and other intraocular media relative to that of a standard observer, d macula is the deviation of the mean value of the macular pigment optical density spectrum relative to the standard observer;

[0141] d L d M dS are the deviations of the peak values ​​of the optical density spectra of the L, M, and S cone cells relative to the standard observer, s L 、s M 、s S are the deviations of the peak wavelengths of the optical density spectra of the L, M, and S cone pigments relative to the standard observer;

[0142] The normalized LMS cone spectral response function in quanta is:

[0143]

[0144] in, are the normalized LMS cone spectral response functions of L, M, and S cone cells in quantum units, α l (λ), α m (λ), α s (λ) are the absorption spectra of the pigments of L, M, and S cone cells, respectively, and D macula (λ) is the optical density spectrum of macular pigment, D ocul (λ) is the optical density spectrum of the lens and other intraocular media;

[0145] The normalized LMS cone spectral response function in energy units is:

[0146]

[0147] in, are the normalized LMS cone spectral response functions of L, M, and S cone cells in energy units, respectively, and max[·] represents the calculated peak value.

[0148] The color confusion tristimulus values ​​(X ti ,Y ti ,Z ti ), the background point tristimulus value of the standard observer (X 0 ,Y 0 ,Z 0 ), changing color confusion tristimulus value (X' ti ,Y' ti ,Z' ti ) and the changing background point tristimulus value (X' 0 ,Y' 0 ,Z' 0 ) are collectively referred to as tristimulus values ​​(X, Y, Z), and the tristimulus values ​​(X, Y, Z) are calculated using the following formula:

[0149]

[0150] Where SPD(λ) is the chromaticity coordinate of the background color (u 0 ,v0 ) corresponds to the spectral SPD 0 (λ), or the MacAdam color confusion threshold chromaticity coordinate (u ti ,v ti ) corresponds to the spectral SPD ti (λ).

[0151] In S34, the color confusion tristimulus values ​​(X ti ,Y ti ,Z ti ) and the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 ,include:

[0152] S341. Calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ), the background point tristimulus value of the standard observer (X 0 ,Y 0 ,Z 0 ) in the CIE 1976 L*u*v* chromaticity coordinate system L*, a*, b* and

[0153]

[0154] Among them, (X, Y, Z) represents the color confusion tristimulus value of the standard observer (X ti ,Y ti ,Z ti ) or the background point tristimulus value of the standard observer (X 0 ,Y 0 ,Z 0 ), (X n ,Y n ,Z n ) represents the tristimulus value of D65 light source, X n =95.047, Y n =100.000, Z n =108.883;

[0155] S342. Calculate the following intermediate variables:

[0156]

[0157] in, is the arithmetic mean of the chroma values ​​of the background point and the threshold point;

[0158] S343. Calculate the following intermediate variables:

[0159]

[0160] in, is the color confusion tristimulus value of the standard observer (X ti ,Y ti ,Z ti )L*, C′ in the CIE 1976 L*u*v* chromaticity coordinate system 1 is the color confusion tristimulus value (X ti ,Y ti ,Z ti ) Calculated C', h' 1 is the color confusion tristimulus value (X ti ,Y ti ,Z ti ) calculated h';

[0161] is the background point tristimulus value of the standard observer (X 0 ,Y 0 ,Z 0 )L*, C′ in the CIE 1976 L*u*v* chromaticity coordinate system 2 is the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Calculated C', h' 2 is the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) calculated h';

[0162] S344, calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ) and the background point tristimulus value (X 0 ,Y 0 ,Z 0 ) Color difference ΔE defined by CIE DE 2000 color difference formula 00 :

[0163]

[0164] in,

[0165]

[0166] R T = -sin(2·Δθ)·RC

[0167]

[0168] for and The average value of C′ 1 and C′ 2 The average value of h′ 1 and h′ 2 The average value of .

[0169] In the technical solution of the present application, the LMS cone spectral response curves are divided into 6 categories according to the different types of color vision defects of the subjects. The value ranges of the parameters in the LMS cone spectral response mathematical model lms(λ,j) corresponding to each type of LMS cone spectral response curve are shown in the following table:

[0170] Table 1 Parameter ranges of various LMS cone spectral response curves

[0171]

[0172]

[0173] In the above table, the value marked with std indicates the standard deviation, and the value without std indicates the absolute value range.

[0174] LMS-a represents the parameter value range of the LMS cone spectral response curve of normal subjects;

[0175] LMS-b represents the range of values ​​of the LMS cone spectral response curve for subjects with dichromatic or trichromatic color vision deficiency;

[0176] LL'S represents the range of values ​​of the LMS cone spectral response curve of subjects with green color vision deficiency (M cells are replaced by mutant L' cells, clinically manifested as green weakness);

[0177] LS represents the range of the LMS cone spectral response curve of subjects with green color vision deficiency (M cell loss, clinically manifested as green blindness);

[0178] MS represents the range of values ​​of the LMS cone spectral response curve of subjects with red color vision deficiency (L cell deficiency, clinically manifested as red blindness);

[0179] M'MS represents the range of values ​​of the LMS cone spectral response curve of subjects with red color vision deficiency (L cells are replaced by mutant M' cells, clinically manifested as red weakness);

[0180] LMS-c represents the range of values ​​of the LMS cone spectral response curve for subjects with color vision defects caused by acquired diseases such as lens or macular degeneration.

[0181] Figures 9 to 12 They are the LMS cone spectral response curves of normal subjects, subjects with red color vision deficiency, subjects with green color vision deficiency, and subjects with blue color vision deficiency, respectively. Among them, 9 is the L cone spectral response curve of the standard observer, 10 is the M cone spectral response curve of the standard observer, 11 is the S cone spectral response curve of the standard observer, 12 is the L cone spectral response curve of the subject, 13 is the M cone spectral response curve of the subject, and 14 is the S cone spectral response curve of the subject.

[0182] Based on the above-disclosed method for quantitatively detecting color vision defects of human eyes, the technical solution of the present application also discloses a device for detecting color vision defects of human eyes, such as Figure 8 As shown, it includes a computer, a colorimeter, a spectrometer, a display and an input device;

[0183] A colorimeter is used to correct the color difference of the display to within the allowable range. When correcting the color difference, the colorimeter generates a source color calibration pattern and sends it to the display, and at the same time receives the color calibration pattern displayed by the display, compares the color calibration pattern with the source color calibration pattern, calculates the color calibration file, and sends it to the computer;

[0184] A spectrometer is used to collect spectral information of the test pattern. When the chromaticity of the display pattern is the same as the chromaticity of the D65 light source, the display brightness is calculated based on the spectral information collected by the spectrometer to adjust the display brightness during the test;

[0185] A display, used to display a calibration pattern to the colorimeter, transmit spectral information of the detection pattern to the spectrometer, display the detection pattern and the detection result to the subject, and display operation interaction information;

[0186] The computer controls the operation of the colorimeter, the spectrometer and the display, and generates the detection pattern and sends it to the display for display;

[0187] An input device is used for the subject to provide feedback on the direction of movement of the dynamic sight mark during or after the movement of the dynamic sight mark. The subject can press the direction keys of the input device such as a keyboard to select left (such as the number key "7"), right (such as the number key "9") or unrecognizable (such as the number key "0") to provide feedback on the direction of movement of the dynamic sight mark.

[0188] Before the device leaves the factory, the colorimeter can be removed after the calibration file is generated and saved. As the device is used for a long time, when the color difference of the display changes significantly, the colorimeter can be installed. This can save the cost of mass production and reduce the instrument procurement cost for end users.

[0189] When the color difference of the display changes slowly, or the LED light source used by the display decays slowly, the spectral information of the detection pattern can also be collected before the device leaves the factory, stored in a computer, and the spectrometer can be removed to save production and procurement costs.

[0190] In the technical solution of the present application, the device provided to the end user can be a full-state device consisting of a computer, a colorimeter, a spectrometer and a display, or it can be a display and a computer after color difference correction and spectrum pre-collection. In simpler cases, it can also be just a laptop computer, tablet computer or mobile phone, which can greatly simplify the hardware structure of the device and lower the usage threshold for end users.

[0191] For some scenarios that only require rapid qualitative testing, chromatic aberration correction and spectral pre-collection can also be omitted. End users can deploy an integrated software package developed based on the qualitative detection method mentioned in this application (such as S1 and S2) on their existing devices to achieve rapid screening of the type of color vision deficiency of the subject. Although the LMS cone spectral response curve of the subject cannot be quantitatively calculated, the qualitative test results are also sufficiently reliable. The subject can go to a professional medical institution for further testing when the qualitative test results show a problem, and almost "cost-free" testing can be achieved.

[0192] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for quantitatively detecting color vision defects of human eyes, characterized in that: The following steps are involved: S1. Using the detection pattern that changes alternately in two dimensions of time and space, combined with the dichotomy chromaticity threshold interactive feedback test method, gradually approaching the color discrimination limit of the subject, and obtaining the subject's MacAdam color confusion ellipse; S2, inputting the subject's MacAdam color confusion ellipse into the trained color vision deficiency type recognition model to obtain the subject's color vision deficiency type; S3. Determine the changed LMS cone spectral response curve according to the type of color vision deficiency of the subject, and obtain the LMS cone spectral response curve of the subject by combining the LMS cone spectral response curve of the standard observer.

2. The method for quantitatively detecting color vision deficiency of human eyes according to claim 1, characterized in that: S1 uses a detection pattern that alternates in two dimensions of time and space, combined with a binary chromaticity threshold interactive feedback test method, to gradually approach the subject's color discrimination limit and obtain the subject's MacAdam color confusion ellipse, including: The angles determined by the red, green, and blue confusion lines of the background color chromaticity coordinates of the detection pattern are used as sampling angles to obtain the six points that best reflect the color vision problem to approximately fit the subject's MacAdam color confusion ellipse; The number of steps n corresponding to the starting minimum boundary of the binary division in each sampling direction min is 0.0001 chromaticity coordinate unit, that is, n min =1, the number of steps corresponding to the starting highest boundary n max is the Euclidean distance between the sRGB boundary and the background coordinate point in that direction, and the step size Δl is 0.0001 chromaticity coordinate unit; The number of steps n corresponding to the lowest boundary is continuously updated during the process of the subject's feedback on the direction of the dynamic sight mark movement. min The number of steps n corresponding to the highest boundary max , when |n max -n min |≤1, and the subject cannot identify the direction of motion of the dynamic sight mark for multiple consecutive times, the direction test is terminated, and the step length corresponding to the highest boundary obtained is n max It is defined as the chromaticity discrimination threshold of the subject at the sampling angle; Among them, the red, green and blue confusion lines are defined as straight lines determined by the intersection points of all red, green and blue confusion lines and the background color chromaticity coordinates. In the CIE 1976L*u*v* chromaticity coordinate system, the intersection point of all red confusion lines is (0.678, 0.501), the intersection point of all green confusion lines is (-1.217, 0.782), and the intersection point of all blue confusion lines is (0.257, 0). When the background point is the chromaticity corresponding to the D65 standard light source, that is, the background color chromaticity coordinates are (0.1978, 0.4684), the angles of the red, green and blue confusion lines relative to the positive direction of the u' coordinate axis are 3.8838° or 183.8838°, -12.4979° or 167.5021°, and -82.7967° or 97.2033°, respectively, and there are 6 intersection points with the MacAdam color confusion ellipse.

3. The method for quantitatively detecting color vision deficiency of human eyes according to claim 2, characterized in that: The detection pattern replaces the static sight mark with a moving dynamic sight mark, converts the color vision detection into direction feedback, and avoids the loophole of the conventional detection method that the subject memorizes the detection pattern after training from the detection mechanism; Random brightness noise is added to the background color and the sight mark color of the detection pattern to avoid the loophole that the subject uses the brightness information in the detection pattern to identify the sight mark, thereby improving the reliability of the test result; The dynamic sight mark moves from the lower right corner to the upper left corner of the background image, or from the lower left corner to the upper right corner. The subject provides feedback on the movement direction of the dynamic sight mark through an input device during or after the movement of the dynamic sight mark.

4. The method for quantitatively detecting color vision deficiency of human eyes according to claim 3, characterized in that: The background color chromaticity coordinates (u0, v0) of the detection pattern are set to any point in the CIE1976L*u*v* chromaticity coordinate system, and the average brightness value of all small squares in the background image is 24cd / m 2 , the brightness noise is ±45%, and the background image covers the field of view of the human eye of more than 10°*10°; The visual standard chromaticity coordinates (u t ,v t ) is calculated using the following formula: Where ΔL is the Euclidean distance of the current test, and θ is the angle of the current test; All the small squares in the dynamic sight mark correspond to the same brightness value, the brightness mean and brightness noise of all the small squares in the dynamic sight mark are the same as the background image, and the dynamic sight mark covers a 2°*2° field of view of the human eye.

5. The method for quantitatively detecting color vision deficiency of human eyes according to claim 1, characterized in that: In S2, the subject's MacAdam color confusion ellipse is input into the trained color vision defect type recognition model to obtain the subject's color vision defect type, including: S21, collecting MacAdam color confusion ellipses and corresponding color vision deficiency types of a large number of subjects, annotating the MacAdam color confusion ellipses using the color vision deficiency types, constructing a historical data set, and dividing the historical data set into a training set, a validation set, and a test set according to a preset ratio; S22, setting a loss function and an optimizer for the color vision deficiency type recognition model; S23, inputting the training set into the color vision deficiency type recognition model for model training; S24, calculating the loss value based on the loss function, and the optimizer updates the model parameters according to the loss value and the network gradient information; S25. If the loss value is less than the preset threshold, the model training ends, and the current color vision deficiency type recognition model is the trained color vision deficiency type recognition model. Otherwise, the process returns to S23 and continues to use the training set for model training. S26. Input the validation set into the trained color vision deficiency type recognition model, evaluate the generalization ability of the model by observing the performance on the validation set, and tune the hyperparameters and structure of the model; S27. Input the test set into the optimized color vision deficiency type recognition model to evaluate the model performance.

6. The method for quantitatively detecting color vision deficiency of human eyes according to claim 1, characterized in that: In S3, a changed LMS cone spectral response curve is determined according to the type of color vision deficiency of the subject, and the LMS cone spectral response curve of the subject is obtained by combining the LMS cone spectral response curve of the standard observer, including: S31, test to obtain the subject's MacAdam color confusion threshold chromaticity coordinates (u ti ,v ti ), i∈[1,6]; S32, test or retrieve the background color chromaticity coordinates (u0, v0), MacAdam color confusion threshold chromaticity coordinates (u ti ,v ti ) corresponding to the spectra SPD0(λ), SPD ti (λ); S33, using the LMS cone spectral response curve lms of the standard observer standard (λ) Calculate the color confusion tristimulus value (X) of the standard observer in the CIE 1931XYZ tristimulus value coordinate system ti ,Y ti ,Z ti ), background point tristimulus values ​​(X0, Y0, Z0) of the standard observer; S34. Calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ) and the background point tristimulus value (X0, Y0, Z0) of the standard observer, the color difference ΔE defined by the CIE DE 2000 color difference formula 00 ; S35. Determine the LMS cone spectral response curve lms according to the type of color vision deficiency of the subject calculate (λ), using the varying LMS cone spectral response curve lms calculate (λ) Calculate the color confusion tristimulus value (X') in the CIE 1931XYZ tristimulus value coordinate system ti ,Y' ti ,Z' ti ), the changing background point tristimulus values ​​(X'0, Y'0, Z'0); S36, calculate the changed color confusion tristimulus value (X' ti ,Y' ti ,Z' ti ) and the changing background point tristimulus value (X'0, Y'0, Z'0) defined by the CIE DE 2000 color difference formula ΔE 00 '; S37, if ΔE 00 '<ε, then the LMS cone spectral response curve lms is continuously optimized. calculate (λ) parameter, so that |ΔE 00 '-ΔE 00 |<ζ, if ΔE 00 '≥ε or |ΔE 00 '-ΔE 00 |≥ζ, then return to S35; S38, when |ΔE 00 '-ΔE 00 When |<ζ, the changing LMS cone spectral response curve lms can be obtained. calculate The parameters of (λ) are used to obtain the LMS cone spectral response curve of the subject; Among them, the LMS cone spectral response curve of the standard observer is lms standard (λ) is the cone spectral response curve of the standard observer specified in CIE 170-1-2006 obtained when the subject age a=32, the field angle v=2°, and other parameters are all 0 in the LMS cone spectral response mathematical model; Variation of the LMS cone spectral response curve lms calculate (λ) is obtained by taking random values ​​of parameters in the mathematical model of LMS cone spectral response, and the range of values ​​of each parameter is determined according to the type of color vision deficiency of the subject; ε is the color difference threshold, ζ is the color difference difference threshold, and ε and ζ are determined according to the type of color vision deficiency of the subject.

7. The method for quantitatively detecting color vision deficiency of human eyes according to claim 6, characterized in that: The LMS cone spectral response mathematical model lms(λ,j) is expressed by the following formula: lms(λ,j)=f(λ,j,a,v,d lens ,d macula ,d L ,d M ,d S ,s L ,s M ,s S ); Where λ is the wavelength, j is the cone cell type, j = L, M, S, representing L cone cells, M cone cells, and S cone cells, respectively, a is the age of the subject, v is the field of view, v = 2°, d lens is the deviation of the peak value of the optical density spectrum of the lens and other intraocular media relative to that of a standard observer, d macula is the deviation of the mean value of the macular pigment optical density spectrum relative to the standard observer; d L ,d M ,d S are the deviations of the peak values ​​of the optical density spectra of the L, M, and S cone cells relative to the standard observer, s L 、s M 、s S are the deviations of the peak wavelengths of the optical density spectra of the L, M, and S cone pigments relative to the standard observer; The normalized LMS cone spectral response function in quanta is: in, are the normalized LMS cone spectral response functions of L, M, and S cone cells in quantum units, α l (λ), α m (λ), α s (λ) are the absorption spectra of the pigments of L, M, and S cone cells, respectively, and D macula (λ) is the optical density spectrum of macular pigment, D ocul (λ) is the optical density spectrum of the lens and other intraocular media; The normalized LMS cone spectral response function in energy units is: in, are the normalized LMS cone spectral response functions of L, M, and S cone cells in energy units, respectively, and max[·] represents the calculated peak value.

8. The method for quantitatively detecting color vision deficiency of human eyes according to claim 7, characterized in that: The color confusion tristimulus values ​​(X ti ,Y ti ,Z ti ), the background point tristimulus values ​​of the standard observer (X0, Y0, Z0), the changing color confusion tristimulus values ​​(X' ti ,Y' ti ,Z' ti ) and the changing background point tristimulus values ​​(X'0, Y'0, Z'0) are collectively referred to as tristimulus values ​​(X, Y, Z). The tristimulus values ​​(X, Y, Z) are calculated using the following formula: Where SPD(λ) is the spectrum SPD0(λ) corresponding to the background color chromaticity coordinates (u0, v0), or the MacAdam color confusion threshold chromaticity coordinates (u ti ,v ti ) corresponds to the spectral SPD ti (λ).

9. The method for quantitatively detecting color vision deficiency of human eyes according to claim 8, characterized in that: In S34, the color confusion tristimulus values ​​(X ti ,Y ti ,Z ti ) and the background point tristimulus value (X0, Y0, Z0) of the standard observer, the color difference ΔE defined by the CIEDE 2000 color difference formula 00 ,include: S341. Calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ), the background point tristimulus values ​​(X0, Y0, Z0) of the standard observer in the CIE 1976L*u*v* chromaticity coordinate system L*, a*, b* and Among them, (X, Y, Z) represents the color confusion tristimulus value of the standard observer (X ti ,Y ti ,Z ti ) or the background point tristimulus value (X0, Y0, Z0) of the standard observer, (X n ,Y n ,Z n ) represents the tristimulus value of D65 light source, X n =95.047, Y n =100.000, Z n =108.883; S342. Calculate the following intermediate variables: in, is the arithmetic mean of the chroma values ​​of the background point and the threshold point; S343. Calculate the following intermediate variables: in, is the color confusion tristimulus value of the standard observer (X ti ,Y ti ,Z ti ) L* in the CIE 1976 L*u*v* chromaticity coordinate system, C′1 is the color confusion tristimulus value (X ti ,Y ti ,Z ti ) calculated C', h'1 is the color confusion tristimulus value (X ti ,Y ti ,Z ti ) calculated h'; is the L* of the background point tristimulus values ​​(X0, Y0, Z0) of the standard observer in the CIE 1976L*u*v* chromaticity coordinate system, C′2 is the C' calculated from the background point tristimulus values ​​(X0, Y0, Z0) of the standard observer, and h′2 is the h' calculated from the background point tristimulus values ​​(X0, Y0, Z0) of the standard observer; S344, calculate the color confusion tristimulus value (X ti ,Y ti ,Z ti ) and the background point tristimulus value (X0, Y0, Z0) of the standard observer, the color difference ΔE defined by the CIE DE 2000 color difference formula 00 : in, for and The average value of is the average value of C′1 and C′2, is the average value of h′1 and h′2.

10. A human eye color vision defect detection device, used to execute the human eye color vision defect quantitative detection method according to claim 1, characterized in that: Includes computers, colorimeter, spectrometer, monitor and input devices; A colorimeter is used to correct the color difference of the display to within the allowable range. When correcting the color difference, the colorimeter generates a source color calibration pattern and sends it to the display, and at the same time receives the color calibration pattern displayed by the display, compares the color calibration pattern with the source color calibration pattern, calculates the color calibration file, and sends it to the computer; A spectrometer is used to collect spectral information of the test pattern. When the chromaticity of the display pattern is the same as the chromaticity of the D65 light source, the display brightness is calculated based on the spectral information collected by the spectrometer to adjust the display brightness during the test; A display, used to display a calibration pattern to the colorimeter, transmit spectral information of the detection pattern to the spectrometer, display the detection pattern and the detection result to the subject, and display operation interaction information; The computer controls the operation of the colorimeter, the spectrometer and the display, and generates the detection pattern and sends it to the display for display; An input device used for the subject to provide feedback on the direction of movement of the dynamic sight mark during or after the movement of the dynamic sight mark.

Images