Cone Spectroscopic Sensitivity Estimator

Abstract

To correctly estimate the spectral sensitivity of a conical body of an individual subject.SOLUTION: A conical body spectral sensitivity estimation apparatus 4 comprises: an experimental data acquisition unit 63 which acquires an RGB value of a test color adjusted by a subject P and an RGB value of a reference color displayed on another region 22 of a screen as experimental data of the condition and the like color experiment that makes the subject P adjust the test color such that the test color displayed on one region 21 of the screen of a display device 2 and the reference color corrected by a color filter 23 installed in the other region 22 of the screen become the same color; and an estimation unit 64 which estimates a conical body spectral sensitivity function indicating the spectral sensitivity of a conical body of the subject P on the basis of the experimental data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] This invention relates to a cone spectral sensitivity estimation device. [Background technology]

[0002] It is known that there are individual differences in people's visual characteristics. Visual characteristics are represented by the spectral sensitivity function (color matching function), which is a curve that shows the sensitivity of cones for each wavelength of light. The color that a person perceives can be determined by the spectral sensitivity function and the spectral radiance of the observed object.

[0003] Patent Document 1 discloses a processing unit comprising: a selection means for causing a user to select a color from among several different colors displayed on a display device that is closest to a reference color formed on a light-reflecting medium; a determination means for determining the color to be displayed on the display device next time based on the color selected by the user; and a calculation means for calculating the user's visual characteristics based on the results of repeating the color determination by the determination means and the selection by the user multiple times. When the user repeatedly selects from a combination of two or more colors a certain number of times or more, the calculation means calculates the user's visual characteristics based on the colors included in the combination. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2016-134683 [Overview of the project] [Problems that the invention aims to solve]

[0005] The processing unit disclosed in Patent Document 1 allows a subject to select a color from a display that is close to a reference color chart in a color matching experiment. In this case, the processing unit disclosed in Patent Document 1 displays a measurement color chart, which is the light source color, on a self-illuminating display relative to a printed reference color. As a result, a difference in color gamut or brightness occurs between the reference color and the measurement color chart in the processing unit disclosed in Patent Document 1, making it difficult to perform a color matching experiment correctly. Therefore, it is difficult to accurately estimate the spectral sensitivity function of the subject's cones from the results of the color matching experiment using the processing unit disclosed in Patent Document 1.

[0006] This invention has been made in view of the above, and aims to accurately estimate the spectral sensitivity of an individual subject's cone cells. [Means for solving the problem]

[0007] To solve the above problems, the present invention provides a cone spectral sensitivity estimation device comprising: an experimental data acquisition unit that acquires the RGB values ​​of the test color adjusted by the subject and the RGB values ​​of the reference color displayed in the other area as experimental data for a conditional color matching experiment in which the subject adjusts the test color so that the test color displayed in one area of ​​the display device screen and the reference color corrected by a color filter installed in another area of ​​the screen become the same color; and an estimation unit that estimates a cone spectral sensitivity function indicating the spectral sensitivity of the subject's cones based on the experimental data.

[0008] The conditional color matching experiment of the present invention can be performed using a simple method of placing a color filter in another area of ​​the display device screen, and can be performed under experimental conditions in which no difference in color gamut or brightness occurs between the reference color and the test color. Furthermore, the cone spectral sensitivity estimation device of the present invention can estimate the cone spectral sensitivity function of a subject based on experimental data obtained from such a simple and accurate conditional color matching experiment. Thus, the cone spectral sensitivity estimation device of the present invention can easily and accurately estimate the cone spectral sensitivity of an individual subject.

[0009] In a more preferred embodiment, the cone spectral sensitivity function is expressed using parameters relating to the subject's visual characteristics, the cone response value to the reference color stimulus is expressed using the spectral radiance of the reference color stimulus and the cone spectral sensitivity function, the cone response value to the test color stimulus is expressed using the spectral radiance of the test color stimulus and the cone spectral sensitivity function, and the estimation unit estimates the cone spectral sensitivity function by searching for parameters that minimize the mean squared error of the relative error between the cone response value to the reference color stimulus and the cone response value to the test color stimulus.

[0010] In this embodiment, the cone spectral sensitivity estimation device of the present invention can avoid the situation in which the cone spectral sensitivity function converges to zero in the optimization calculation that minimizes the mean square error described above, and can also minimize the influence of the reference stimulus on the estimation accuracy of the cone spectral sensitivity function. Therefore, the cone spectral sensitivity estimation device of the present invention can estimate the cone spectral sensitivity function with greater accuracy. Thus, the cone spectral sensitivity estimation device of the present invention can estimate the cone spectral sensitivity of an individual subject with greater accuracy.

[0011] In a more preferred embodiment, the test color is represented by juxtaposed additive color mixing, in which different colors are displayed in adjacent groups of pixels within the same region.

[0012] This embodiment makes it possible to easily achieve high-gradation test colors without having to perform conditional color matching experiments using a special display device capable of high-gradation display. Therefore, the cone spectral sensitivity estimation device of the present invention can accurately estimate the cone spectral sensitivity function of a subject, even if the subject has cones that can perceive subtle differences in color. Therefore, the cone spectral sensitivity estimation device of the present invention can estimate the cone spectral sensitivity of an individual subject even more simply and accurately.

[0013] In a more preferred embodiment, the estimation unit searches for the parameters using differential evolution.

[0014] In this embodiment, the estimation unit of the present invention can search for parameters using a relatively simple algorithm, thereby enabling the easy estimation of the cone spectral sensitivity function. Therefore, the cone spectral sensitivity estimation device of the present invention can further easily estimate the cone spectral sensitivity of an individual subject.

[0015] In a more preferred embodiment, the experimental data acquisition unit acquires a plurality of experimental data obtained by performing multiple condition matching experiments with a plurality of color filters having different spectral transmittances, and the estimation unit estimates the cone spectral sensitivity function based on the plurality of experimental data.

[0016] In this embodiment, the cone spectral sensitivity estimation device of the present invention can estimate the cone spectral sensitivity function using the cone response values ​​of a subject to color stimuli having various wavelength ranges, thereby enabling a more accurate estimation of the subject's cone spectral sensitivity function. Furthermore, the cone spectral sensitivity estimation device of the present invention can suppress estimation errors of the cone spectral sensitivity function even if there is variability in the experimental data of the conditional color matching experiment. Therefore, the cone spectral sensitivity estimation device of the present invention can more accurately estimate the cone spectral sensitivity of an individual subject. [Effects of the Invention]

[0017] According to the present invention, the spectral sensitivity of an individual subject's cones can be accurately estimated. [Brief explanation of the drawing]

[0018] [Figure 1] This figure shows the configuration of a cone spectral sensitivity estimation system equipped with the cone spectral sensitivity estimation device of this embodiment. [Figure 2] A flowchart illustrating the procedure for a conditional color matching experiment. [Figure 3] Figure 3(a) illustrates the conditional color matching experiment, and Figure 3(b) shows the spectral transmittance of each color filter. [Figure 4] A diagram illustrating the enhancement of test colors to higher tonality. [Figure 5]A flowchart illustrating the process of estimating the cone spectral sensitivity function using a cone spectral sensitivity estimation device. [Figure 6] Figure 6(a) shows an example of the results of estimating the cone spectral sensitivity function of a subject using a cone spectral sensitivity estimation device, and Figure 6(b) shows an example of estimation error in the cone spectral sensitivity function due to variability in experimental data. [Modes for carrying out the invention]

[0019] Embodiments of the present invention will be described below with reference to the drawings. Components denoted by the same reference numerals in each embodiment have the same function in each embodiment unless otherwise specified, and their description will be omitted.

[0020] [Configuration of the cone spectroscopic sensitivity estimation system] The configuration of the cone spectral sensitivity estimation system 1 of this embodiment will be explained using Figure 1.

[0021] Figure 1 shows the configuration of the cone spectral sensitivity estimation system 1 equipped with the cone spectral sensitivity estimation device 4 of this embodiment.

[0022] The cone spectral sensitivity estimation system 1 is a system that estimates the spectral sensitivity of subject P's cones as a visual characteristic of subject P. In particular, the cone spectral sensitivity estimation system 1 is a system that estimates the cone spectral sensitivity function, which indicates the spectral sensitivity of subject P's cones, based on experimental data, which are experimental results of a conditional color matching experiment on subject P. The cone spectral sensitivity estimation system 1 comprises a display device 2, an input device 3, and a cone spectral sensitivity estimation device 4.

[0023] Display device 2 consists of a display that shows various colors to subject P in order to conduct a conditional color matching experiment. Input device 3 consists of a general-purpose input device such as a mouse or keyboard that inputs information through subject P's operation during the conditional color matching experiment. Input device 3 may also consist of a dedicated input device such as a slider that is used for subject P's operation during the conditional color matching experiment.

[0024] The cone spectral sensitivity estimation device 4 consists of a computer system that estimates the spectral sensitivity of the cones of a subject P. The cone spectral sensitivity estimation device 4 comprises a storage device 5 and an arithmetic processing unit 6. The storage device 5 consists of an SSD or HDD, etc. The storage device 5 stores various data used in the processing of the arithmetic processing unit 6. The arithmetic processing unit 6 consists of a CPU, ROM, RAM, etc. The arithmetic processing unit 6 realizes the various functions of the cone spectral sensitivity estimation device 4 by having the CPU execute a program stored in the ROM.

[0025] The arithmetic processing unit 6 includes a display control unit 61 that controls the display of the display device 2 in a conditional color matching experiment. Furthermore, the arithmetic processing unit 6 includes an experimental information acquisition unit 62 that acquires experimental information indicating the experimental conditions of the conditional color matching experiment, and an experimental data acquisition unit 63 that acquires experimental data of the conditional color matching experiment. Furthermore, the arithmetic processing unit 6 includes an estimation unit 64 that estimates a cone spectral sensitivity function indicating the spectral sensitivity of the cone of subject P based on the experimental data acquired by the experimental data acquisition unit 63.

[0026] [Conditional Color Matching Experiment] The color matching experiment under the conditions of this embodiment will be explained using Figures 2 to 4.

[0027] Figure 2 is a flowchart showing the flow of a conditional color matching experiment. Figure 3(a) is a diagram illustrating the conditional color matching experiment. Figure 3(b) is a diagram showing the spectral transmittance of each color filter 23. Figure 4 is a diagram illustrating the high-gradation enhancement of the test color.

[0028] In the conditional color matching experiment of this embodiment, the test color displayed in one area 21 of the display device 2's screen (the right side of the screen in Figure 3(a)) is matched to the reference color corrected by the color filter 23 installed in another area 22 of the display device 2's screen (the left side of the screen in Figure 3(a)) by having the subject P adjust the test color so that the test color is the same as the reference color.

[0029] Specifically, in the conditional color matching experiment of this embodiment, the experiment manager or subject P places the color filter 23 in another area 22 of the display device 2 screen (step S1 in Figure 2). The color filter 23 is composed of multiple color filters 23 with different spectral transmittances. For example, as shown in Figure 3(b), the color filter 23 is a color filter in which the color (corrected color) after achromatic light has passed through the filter is a light shade of blue, green, pink, yellow-green, red, or purple. That is, the reference color corrected by achromatic light passing through the color filter 23 is blue, green, pink, yellow-green, red, or purple.

[0030] Next, in the conditional color matching experiment of this embodiment, the display control unit 61 of the cone spectral sensitivity estimation device 4 performs initial display control of one area 21 and the other area 22 of the display device 2 screen (step S2 in Figure 2). The test color, which will be the light source color, is displayed in one area 21 of the screen. The color initially displayed as the test color may be a predetermined achromatic or chromatic color. The reference color before correction by the color filter 23 is displayed in the other area 22 of the screen. The reference color displayed in the other area 22 of the screen remains unchanged until the color filter 23 is replaced.

[0031] Here, the test color is represented by juxtaposition additive color mixing, in which different colors are displayed in adjacent groups of pixels within a region 21 of the screen. That is, as shown in Figure 3(a), the display control unit 61 displays the test color by dividing two or more colors into a checkerboard pattern. For example, suppose that a region 21 of the screen is composed of 6x6 pixels, as shown in Figure 4. In Figure 4, "0" or "1" in each pixel indicates that the light emission of each pixel is "OFF" or "ON". The light emission intensity of each pixel can be set in the range of 0 to 255. The display control unit 61 divides the entire 6x6 pixels into 9 parts to create groups of 2x2 pixels. For each created group of pixels, the display control unit 61 performs 2 8=256 gradations of color are set. In this case, 255 × 9 + 1 = 2296 different colors can be displayed in one area 21 of the screen. In reality, different colors are displayed for each pixel group, but to subject P observing from a reasonable distance from the display device 2, the different colors displayed for each pixel group are mixed and appear as a uniform color. In this way, the test color is represented by juxtaposed additive color mixing. Note that in Figure 4, one area 21 of the screen may be composed of 3 × 3 pixels or 12 × 12 pixels, and is not particularly limited. The colors displayed for each pixel group may also be set within a range in which subject P does not perceive a discontinuity of color, and are not particularly limited.

[0032] Next, in the conditional color matching experiment of this embodiment, the display control unit 61 of the cone spectral sensitivity estimation device 4 performs display control to display a message on the display device 2 prompting subject P to perform color matching. Subject P adjusts the test color so that the test color displayed in one area 21 of the screen matches the reference color corrected by the color filter 23 installed in another area 22 of the screen (step S3 in Figure 2). Subject P can adjust the test color arbitrarily by operating the input device 3. Subject P can adjust the test color by inputting the RGB values ​​of the test color using the input device 3, which is configured as a keyboard. Alternatively, Subject P can adjust the test color by moving the input device 3, which is configured as a slider or the like. The input device 3 notifies the display control unit 61 of the RGB values ​​of the test color adjusted by subject P. The display control unit 61 controls the RGB values ​​of each pixel in one area 21 of the screen according to the RGB values ​​notified from the input device 3. As a result, the test color adjusted by subject P is displayed in one area 21 of the screen. After adjustment by subject P, the display control unit 61 uses the RGB values ​​of the test color adjusted by subject P as input values ​​(R t ,G t ,B t The RGB values ​​of the reference color (reference color before correction by the color filter 23) displayed in the other area 22 of the screen are stored in the storage device 5 as input values ​​(R r ,G r,B r ) is stored in memory device 5.

[0033] Next, in the conditional color matching experiment of this embodiment, the experiment manager or subject P determines whether or not all color filters 23 have been tested (step S4 in Figure 2). If not all color filters 23 have been tested, the color filters 23 are replaced with untested ones (step S1), and the conditional color matching experiment is performed again. If all color filters 23 have been tested, the conditional color matching experiment is terminated. The determination of whether or not all color filters 23 have been tested may be made by the display control unit 61 by checking the RGB values ​​of the test color and reference color stored in the storage device 5. The display control unit 61 may then execute display control to display a message on the display device 2 indicating that the conditional color matching experiment has been completed, or a message prompting the experiment manager or subject P to replace the color filters 23.

[0034] [Estimation process for cone spectral sensitivity function] Using Figure 5, the cone spectral sensitivity function estimation process performed by the cone spectral sensitivity estimation device 4 of this embodiment will be explained.

[0035] Figure 5 is a flowchart showing the flow of the cone spectral sensitivity function estimation process performed by the cone spectral sensitivity estimation device 4.

[0036] First, the experimental information acquisition unit 62 of the cone spectral sensitivity estimation device 4 acquires experimental information (step S11 in Figure 5). The experimental information includes the age a [years] of subject P in the conditional color matching experiment.

[0037] Subsequently, the experimental data acquisition unit 63 of the cone spectral sensitivity estimation device 4 acquires the experimental data of the conditional metamerism experiment (step S12 in FIG. 5). Specifically, for each conditional metamerism experiment, the experimental data acquisition unit 63 acquires from the storage device 5 the input values (Rt, Gt, Bt), which are the RGB values of the test color adjusted by the subject P, and the input values (Rr, Gr, Br), which are the RGB values of the reference color displayed in the other area 22 of the screen. When the conditional metamerism experiment is performed multiple times with a plurality of color filters 23 having different spectral transmittances, the experimental data acquisition unit 63 acquires a plurality of experimental data.

[0038] Subsequently, the estimation unit 64 of the cone spectral sensitivity estimation device 4 estimates the cone spectral sensitivity function of the subject P based on the acquired experimental information and experimental data. Specifically, the estimation unit 64 performs the processes shown in steps S13 to S16 of FIG. 5. When the conditional metamerism experiment is performed multiple times, the estimation unit 64 estimates the cone spectral sensitivity function of the subject P based on the plurality of experimental data.

[0039] First, the estimation unit 64 identifies the parameters regarding the visual characteristics of the subject P (step S13 in FIG. 5). Specifically, the estimation unit 64 identifies ten parameters including the age a [years] of the subject P, the stimulus size ν [deg], the deviation d lens [%] of the optical density of the ocular optical media, the deviation d macula [%] of the macular optical density, the deviation d L [%], d M [%], d S [%] of the maximum optical density of each cone, and the wavelength direction offsets s L [nm], s M [nm], s S [nm]. Note that the following cone spectral sensitivity function is represented using the parameters regarding the visual characteristics of the subject P.

[0040] The stimulus size ν is the size of the color stimulus as seen by the subject P in the conditional metamerismism experiment. The stimulus size ν is represented by the visual angle [deg]. The stimulus size ν is not explicitly input as the size of the color stimulus set as an experimental condition, but is set as a parameter regarding the visual characteristics of the subject P.

[0041] Deviation d of optical density of the ocular optical medium lens This is the deviation of the optical density of the ocular optical medium of subject P from the optical density of the ocular optical medium of a standard observer, which is used when defining standard cone spectral sensitivity. This standard observer may be a standard observer designated by the CIE (International Commission on Illumination).

[0042] Macular optical density deviation d macula This is the deviation of subject P's macular optical density from that of the standard observer.

[0043] The deviation d of the maximum optical density of each cone cell. L d M d S This is the deviation of the maximum optical density of each cone in subject P from the maximum optical density of each cone in the standard observer.

[0044] Shift of spectral absorbance in the wavelength direction s L ,s M ,s S s is the chromatic deviation of the spectral absorbance of the visual pigment in each cone cell of subject P from the spectral absorbance of the visual pigment in each cone cell of a standard observer. The spectral sensitivity of each of the three cones, L, M, and S, is a function of the wavelength of light, and its waveform can be roughly bell-shaped. L ,s M ,s S This can directly affect the wavelength of the maximum spectral sensitivity of each of the three cones. Therefore, the estimation unit 64 considers the wavelength shift of spectral absorbance s, which has the greatest influence on the estimation of the cone spectral sensitivity function among the 10 parameters mentioned above. L ,s M ,s S At least identify them.

[0045] Next, the estimation unit 64 creates a visual model representing the cone spectral sensitivity function of subject P using the identified parameters (step S14 in Figure 5). Specifically, the estimation unit 64 creates Equation 1 as a visual model representing the cone spectral sensitivity function of subject P. λ represents wavelength. Spectral sensitivity function l q(λ),m q (λ),s q (λ) is the spectral absorbance α, as shown in Equation 1. j (λ)(j=L,M,S), macular optical density D macula (λ), ocular optical medium optical density D lens It is represented by (λ).

[0046]

number

[0047] If the age of subject P is 60 years or younger, then according to formula 2, the average optical density of the ocular optical medium D lens,ave (λ) is defined. If the age a of subject P is greater than 60 years, then the average optical density of the ocular optical medium D is given by formula 3. lens,ave (λ) is defined. Also, as shown in equation 4, the deviation d of the optical density of the ocular optical medium. lens This will be introduced to express individual differences.

[0048]

number

[0049]

number

[0050]

number

[0051] Macular optical density D macula (λ) is the average macular optical density D, as shown in equations 5 and 6. macula,ave The influence of (λ) and stimulus size ν on the deviation of macular optical density d macula This will be introduced to express individual differences.

[0052]

number

[0053]

number

[0054] Wavelength shift s j The spectral absorbance A of each cone at low optical density reflects this. shift,j (λ) is expressed by equation 8, where ν is the stimulus size and d is the deviation of the maximum optical density of each cone. L d M d S Using this, the maximum value D of the optical density of each cone reflects individual differences. max,j This is expressed by equation 9. The spectral absorbance α of each cone j (λ) is the spectral absorbance A of each cone at low optical density. shift,j (λ) and the maximum optical density D of each cone max,j Using these, it is expressed by equation 7.

[0055]

number

[0056]

number

[0057]

number

[0058] Based on the above, the visual model representing the cone spectral sensitivity function of subject P is given by subject P's age a [years], stimulus size ν [deg], and deviation d of the optical density of the ocular optical medium. lens [%], deviation of macular optical density d macula [%], deviation d of the maximum optical density of each cone L [%],d M [%],d S [%], and the wavelength shift of spectral absorbance s L [nm],s M [nm],s SBy using a total of 10 parameters in [nm], individual differences in cone spectral sensitivity can be expressed.

[0059] Next, the estimation unit 64 calculates the cone response values ​​for each reference color stimulus (also referred to as the "reference stimulus") and the test color stimulus (also referred to as the "test stimulus") (step S15 in Figure 5). The cone response values ​​for each reference stimulus and the test stimulus are calculated based on the spectral radiance P of the reference stimulus and the test stimulus. r (λ), P t It is expressed using (λ) and the cone spectral sensitivity functions l(λ), m(λ), and s(λ) shown in Equation 1.

[0060] Let r(λ), g(λ), and b(λ) be the spectral radiances when the input values ​​to the RGB channels of the display device 2 are maximized, and let τ(λ) be the spectral transmittance of the color filter 23. Input values ​​(R) in the reference stimulus and test stimulus r ,G r ,B r ),(R t ,G t ,B t Let ) be between 0 and 1. The gamma characteristics of the display device 2 are set to linear (γ=1). In this case, the spectral radiance P of the reference stimulus is r (λ) is expressed by equation 10. Spectral radiance P of the test stimulus t (λ) is expressed by equation 11. Input values ​​(R) in the reference stimulus and test stimulus. r ,G r ,B r ),(R t ,G t ,B t ) is a value stored in the storage device 5 by the display control unit 61, as described above.

[0061]

number

[0062]

number

[0063] The cone response values ​​L, M, and S for the reference stimulus and test stimulus are expressed by equation 12. In a conditional color matching experiment, the equation in equation 13 should hold. However, due to its form, equation 13 cannot be solved analytically using linear algebra. Therefore, the estimation unit 64 needs to estimate the cone spectral sensitivity function by numerically solving equation 13. In this process, the estimation unit 64 uses multiple experimental data obtained from multiple conditional color matching experiments to numerically solve equation 13.

[0064]

number

[0065]

number

[0066] Next, the estimation unit 64 searches for parameters related to the visual characteristics of subject P so as to minimize the mean squared error of the relative error between the cone response value to the reference stimulus and the cone response value to the test stimulus (step S16 in Figure 5). Specifically, the cone response values ​​to the reference stimulus and the test stimulus in the nth conditional color matching experiment performed by subject P are (L r,n M r,n ,S r,n ),(L t,n M t,n ,S t,n ) The error (residual) between the pyramidal response value to the reference stimulus and the pyramidal response value to the test stimulus is ΔL. n =L r,n -L t,n ,ΔM n =M r,n -M t,n ,ΔS n =S r,n -S t,n In this case, if we find the cone spectral sensitivity function that minimizes the error, the obtained cone spectral sensitivity function will be the individual cone spectral sensitivity function of subject P.

[0067] In this embodiment, the estimation unit 64 estimates the cone spectral sensitivity function of the subject P by searching for the above 10 parameters so that the mean square error of the relative error between the cone response value for the reference stimulus and the cone response value for the test stimulus is minimized. The root mean square error RMSE LMS of the relative error between the cone response value for the reference stimulus and the cone response value for the test stimulus is represented by Equation 14. In Equation 14, ΔL n / L r,n , ΔM n / M r,n , ΔS n / S r,n [[ID=​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​When the mean squared error is used as the evaluation function value, if the cone spectral sensitivity function is brought close to zero across all wavelengths, the mean squared error (evaluation function value) will also approach zero, making it highly likely that the optimization calculation will not function effectively. The inventor has actually performed optimization calculations using the mean squared error as the evaluation function value and confirmed that a cone spectral sensitivity function was obtained in which all values ​​are always zero. On the other hand, the mean squared error RMSE of the relative error LMS When using this as the evaluation function value, even if the cone spectral sensitivity function is brought close to zero, the mean squared error RMSE of the relative error remains. LMS Since both the numerator and denominator of the (evaluation function value) approach zero, it is impossible for the cone spectral sensitivity function that minimizes the evaluation function value to converge to a function where all wavelength ranges are zero. For this reason, the estimation unit 64 calculates the mean squared error (RMSE) of the relative error. LMS This is used as the evaluation function value.

[0070] Furthermore, the mean squared error (RMSE) of the relative error LMS Another reason for adopting ΔL as the evaluation function value is to minimize the influence of the reference stimulus on the estimation accuracy. That is, the balance of the response values ​​of the three types of cones (L cones, M cones, S cones) changes depending on the color of the reference stimulus. For example, with a red reference stimulus, the response value of the L cone will be relatively larger than the response value of the M cone or the S cone. Therefore, simply the error ΔL n ,ΔM n ,ΔS n If the mean squared error (RMSE) is used as the evaluation function value, and the reference stimulus used in the conditional color matching experiment is biased towards red, the error in the response value of the L cone in the mean squared error will have a relatively large weight, which may lead to an inability to properly estimate the cone spectral sensitivity function of the M cone or S cone. The same applies if the reference stimulus is biased towards green or blue. On the other hand, the mean squared error (RMSE) of the relative error... LMSWhen using this as the evaluation function value, it is thought that the cone spectral sensitivity functions of the three types of cones can be estimated in a balanced manner without being affected by the color bias of the reference stimulus. In other words, the response values ​​between LMS cones can be equalized, and the possibility that the spectral sensitivity function of the cone with the largest response value will dominate the sum of squared errors can be eliminated. Moreover, the error between the cone response value to the reference stimulus and the cone response value to the test stimulus in a conditional color matching experiment is proportional to the magnitude of the color discrimination threshold (the smallest color difference that can distinguish between colors). It is known that this discrimination threshold follows Weber's law, which is proportional to the stimulus intensity. Therefore, from the perspective of the characteristics of the experimental data used in the optimization calculation, the mean squared error RMSE of the relative error is also important. LMS It is appropriate to adopt as the evaluation function value. For this reason, the estimation unit 64 uses the mean squared error RMSE of the relative error. LMS This is used as the evaluation function value.

[0071] Mean squared error RMSE expressed by formula 14 LMS To find the parameters that minimize the mean squared error (RMSE), it is necessary to apply some kind of nonlinear optimization method. The mean squared error (RMSE) is expressed by Equation 14. LMS The model is complex, and it is not possible to calculate the derivative of the cone spectral sensitivity function. Therefore, the mean squared error RMSE, expressed by Equation 14, is not applicable. LMS To find the parameters that minimize [the specified value], gradient methods such as the steepest descent method or Newton's method cannot be applied. Furthermore, since the search may result in finding parameters that converge to a local minimum, a global optimization method is preferable as the optimization technique to be applied.

[0072] Therefore, in this embodiment, the estimation unit 64 applies the differential evolution method, which is one of the evolutionary computation algorithms with a relatively simple algorithm, as an optimization method to search for the above parameters. In the minimization of a multivariate function by differential evolution, a large number of vectors (individuals) with the variables as components are generated, and the next generation of individuals is created through mutation and crossover, while the individuals with smaller function values ​​are selectively evolved to finally obtain the minimum value of the function. With this method, the above parameters can be easily determined, and the cone spectral sensitivity function of subject P can be derived.

[0073] The objective here is not to minimize the multivariate function, but rather the mean squared error RMSE. LMS The goal is to find the parameters that minimize [a certain value]. Therefore, a vector with D parameters as components is treated as an individual. The total number of elements in generation G is N. P The i-th element in the group (i=1,2,···,N) P The vector (individual) of x i,G =(x 1i,G ,x 2i,G ,···,x Di,G Let's assume the mutation vector v in generation G+1. i,G+1 This is a vector x randomly selected from the population of generation G. r1,G ,x r2,G ,x r3,G (However, r1, r2, r3 ∈ {i = 1, 2, ..., N P Using}) and the scale coefficient F∈[0,2], it is generated by equation 15.

[0074]

number

[0075] Trial vector u of generation G+1 i,G+1 The j-th parameter u ji,G+1 (j=1,2,···,D) is the parameter x of the original vector ji,G and the parameter v of the mutation vector ji,G+1From this, it is generated by an operation called crossover, as shown in equation 16. At this time, one integer random number j r ∈{1,2,···,D} and j random numbers for each parameter ji Generates a random number ∈[0,1]. ji A parameter such that the crossover rate CR∈[0,1] is less than or equal to a pre-set value, or j r The j-th parameter that matches is the parameter v of the mutation vector. ji,G+1 Replace with the parameter x of the original vector. ji,G Save it.

[0076]

number

[0077] The generated trial vector u i,G+1 and the original vector x i,G Using and , the evaluation value is calculated, and as shown in equation 17, the vector with the lower evaluation value is used as the next-generation vector x i,G+1 This is retained. This process is repeated to evolve the parameters to give the minimum evaluation value. The estimation unit 64 uses the mean squared error RMSE expressed by equation 14. LMS Minimize D = 10 parameters (a, ν, d lens d macula d L d M d S ,s L ,s M ,s S Explore ).

[0078]

number

[0079] Next, the estimation unit 64 derives the cone spectral sensitivity function of subject P from the searched parameters (step S17 in Figure 5). Specifically, the estimation unit 64 uses the 10 searched parameters (a, ν, d lens d macula d Ld M d S ,s L ,s M ,s S Substitute the given values ​​into equations 1 to 9. This allows the estimation unit 64 to derive the cone spectral sensitivity function of subject P.

[0080] [Estimated results of cone spectral sensitivity function] The estimation results of the cone spectral sensitivity function by the cone spectral sensitivity estimation device 4 of this embodiment will be explained using Figures 6(a), 6(b), and Table 1.

[0081] Figure 6(a) shows an example of the results of estimating the cone spectral sensitivity function of subject P using the cone spectral sensitivity estimation device 4. Figure 6(b) shows an example of estimation error in the cone spectral sensitivity function due to variability in experimental data. Table 1 is a table showing an example of the results of exploring parameters related to the visual characteristics of subject P using the cone spectral sensitivity estimation device 4.

[0082] [Table 1]

[0083] In Figures 6(a) and 6(b), the vertical axis CFF (Cone Fundamental Function) represents the value of the cone spectral sensitivity function. The horizontal axis in Figures 6(a) and 6(b) represents wavelength. The solid lines in Figures 6(a) and 6(b) represent the cone spectral sensitivity function estimated by the cone spectral sensitivity estimation device 4. The dashed lines marked with * in Figures 6(a) and 6(b) represent the cone spectral sensitivity function calculated in advance in the simulation, and represent an ideal cone spectral sensitivity function with virtually no variability in the experimental data from the conditional color matching experiment.

[0084] The cone spectral sensitivity function estimation result obtained by the cone spectral sensitivity estimation device 4 is in close agreement with the simulation result showing an ideal cone spectral sensitivity function, as shown in Figure 6(a). In other words, the cone spectral sensitivity estimation device 4 can accurately estimate the cone spectral sensitivity function of subject P. To put it another way, the cone spectral sensitivity estimation device 4 can accurately estimate the spectral sensitivity of an individual subject's cones.

[0085] In conditional color matching experiments, a person observes and matches a reference color and a test color. Experimental data from conditional color matching experiments inevitably contain some errors and variability. These errors and variability can lead to estimation errors in the cone spectral sensitivity function. To suppress these estimation errors, it is effective to conduct multiple conditional color matching experiments under the same conditions.

[0086] Figure 6(a) shows the estimation results of the cone spectral sensitivity function when the variability of the experimental data is small (0.5%) and the number of trials of the conditional color matching experiment is 12. Figure 6(b) shows the estimation results of the cone spectral sensitivity function when the variability of the experimental data is large (2%) and the number of trials of the conditional color matching experiment is 3. In the case of Figure 6(a), the estimation error of the cone spectral sensitivity function is extremely small. In the case of Figure 6(b), the estimation error for the spectral sensitivity function l(λ) of the L cone is large. After various studies, the inventors have confirmed that even with a variability of 1.5% in the experimental data, the estimation error of the cone spectral sensitivity function becomes extremely small if the number of trials of the conditional color matching experiment is 12. Furthermore, the inventors have confirmed that even for subjects P whose cone spectral sensitivity function shape deviates somewhat from the shape of the function shown in Equation 1, the estimation error of the cone spectral sensitivity function can be suppressed to a usable level by performing the conditional color matching experiment multiple times.

[0087] Although embodiments of the present invention have been described in detail above, the present invention is not limited to the embodiments described above, and various modifications can be made without departing from the spirit of the invention as described in the claims. The present invention can be modified by adding the configuration of one embodiment to the configuration of another embodiment, replacing the configuration of one embodiment with that of another embodiment, or deleting a part of the configuration of one embodiment. [Explanation of symbols]

[0088] 2...Display device, 21...One region, 22...Other regions, 23...Color filter, 4...Cone spectral sensitivity estimation device, 63...Experimental data acquisition unit, 64...Estimation unit, P...Subject

Claims

1. An experimental data acquisition unit acquires the RGB values ​​of the test color adjusted by the subject and the RGB values ​​of the reference color displayed in the other area as experimental data for a conditional color matching experiment in which a subject adjusts the test color so that the test color displayed in one area of ​​the display device screen matches the reference color corrected by a color filter installed in another area of ​​the screen. The system includes an estimation unit that estimates a cone spectral sensitivity function representing the spectral sensitivity of the subject's cones based on the experimental data, The estimation unit estimates the cone spectral sensitivity function using the spectral radiance of the test color stimulus corresponding to the RGB value of the test color and the spectral radiance of the reference color stimulus corresponding to the RGB value of the reference color. A cone spectral sensitivity estimation device characterized by the following features.

2. The cone spectral sensitivity function is expressed using parameters relating to the subject's visual characteristics, The response value of the cone to the stimulus of the reference color is expressed using the spectral radiance of the stimulus of the reference color and the spectral sensitivity function of the cone. The response value of the cones to the stimulus of the test color is expressed using the spectral radiance of the stimulus of the test color and the spectral sensitivity function of the cones. The estimation unit estimates the cone spectral sensitivity function by searching for parameters that minimize the mean squared error of the relative error between the response value of the cone to the stimulus of the reference color and the response value of the cone to the stimulus of the test color. The cone spectral sensitivity estimation device according to feature 1.

3. The aforementioned test color is represented by juxtaposed additive color mixing, in which different colors are displayed in adjacent groups of pixels within the aforementioned region. The cone spectral sensitivity estimation device according to feature 1.

4. The estimation unit searches for the parameters using differential evolution. The cone spectral sensitivity estimation device according to feature 2.

5. The experimental data acquisition unit acquires multiple experimental data obtained by performing multiple color matching experiments under the conditions, using multiple color filters with different spectral transmittances. The estimation unit estimates the cone spectral sensitivity function based on a plurality of experimental data. The cone spectral sensitivity estimation device according to feature 1.

Images