A method for evaluating visual noise rejection capability based on contrast modulation grating

Abstract

A visual noise exclusion ability evaluation method based on contrast modulation grating, comprising the following steps: S1, combining the constructed grating noise image; S2, obtaining the contrast modulation grating image with specific noise; S3, the display of the grating is directly opposite to the subject with a set distance, the display displays a general gray background with the brightness being the maximum brightness of the display, and the evaluation is started; S4, after the image appears, the grating is randomly oriented, appears on the screen for a set time and disappears, and then the subject answers the orientation of the grating; when the answer of the subject is consistent with the orientation of the grating, the contrast of the grating in the next test is reduced until the subject cannot judge correctly. The noise information with the same or close spatial frequency as the grating picture is superimposed on the sine grating picture, so that the visual noise information exclusion ability of the subject at the spatial frequency can be more accurately evaluated.

Description

Technical Field

[0001] This invention relates to the technical field of visual images, and more particularly to a method for evaluating the visual noise reduction capability based on a contrast modulation grating. Background Technology

[0002] Visual noise can be understood as "factors that hinder people's sensory organs from understanding the information received from the source." For example, in a black and white image, the brightness distribution that interferes with its reception can be called image noise. Noise can theoretically be defined as "unpredictable random errors that can only be understood using probabilistic and statistical methods." Summary of the Invention

[0003] To evaluate visual noise cancellation capability, this invention proposes a method for evaluating visual noise cancellation capability based on a contrast modulation grating, the specific scheme of which is as follows:

[0004] A method for evaluating the visual noise cancellation capability based on a contrast modulation grating includes the following steps:

[0005] S1. A grating noise image constructed based on a combination of grating size, grating contrast, grating spatial frequency, grating phase, number of pixels per viewing angle of the display, noise level, noise distribution, and noise contrast. The generation function of the grating target is:

[0006]

[0007] Where x, y are the coordinates of the image pixels, with the center of the image as the coordinate (0,0), C is the raster contrast ratio, S is the raster spatial frequency, P is the number of pixels per viewing angle of the display, θ is the raster angle, and γ is the raster phase.

[0008] S2. Obtain a contrast-modulated grating image with specific noise, wherein the noise is represented in the form of square noise dots, and the noise is represented by three factors: the noise pixel size L, and the probability P. n Randomly distributed noise contrast correction value C noice ;

[0009] S3. The display screen showing the raster is facing the subject at a set distance, and the display brightness is set to a gray background at the maximum brightness of the display screen. Start the evaluation.

[0010] S4. After the image appears, the grating is randomly oriented and disappears from the screen after a set time. The subject then answers the orientation of the grating. When the subject's answer matches the grating orientation, the contrast of the grating in the next test is [value missing]. C is the raster contrast of the previous raster image, and R is a number greater than 1. The contrast is reduced until the subject cannot judge correctly. This contrast value is recorded as the subject's visual noise elimination ability score at this spatial frequency and noise contrast.

[0011] Specifically, the orientation of the grating is randomly either horizontal or vertical.

[0012] Specifically, the steps in step S2 to obtain a contrast-modulated grating image with specific noise are as follows:

[0013] S21. Calculate the size L of the noise pixel;

[0014] S22. Divide the raster image into a grid; divide the raster image into equal horizontal and vertical segments with noise pixel size L as the unit;

[0015] S23. Generate noise points according to the noise distribution probability; for each of the above segmented regions, with probability P n The area is determined to be dark; otherwise, it is considered bright. If it is bright, the brightness of the area is increased by C. noise If it is dark, the brightness of that area decreases by C. noise .

[0016] Specifically, the steps for calculating the size L of the noisy pixels are as follows:

[0017] SA211. Calculate the maximum noise pixel width L that matches the grating spatial frequency S. max and round down:

[0018]

[0019] SA212, with L max Using the initial value as an integer, we decrease the value by -1 to find the first integer that can be divided by the image pixel size W, which is the noise pixel size L.

[0020] Specifically, the time in step S4 is set to 150ms.

[0021] The beneficial effect of the present invention is that by superimposing noise information with the same or similar spatial frequency as the grating image onto the sinusoidal grating image, the ability of the subject to eliminate visual noise information at that spatial frequency can be more accurately assessed. Attached Figure Description

[0022] Figure 1 These are before-and-after images of the split.

[0023] Figure 2 These are images showing the effects before and after brightness correction.

[0024] Figure 3 The left image shows a vertical grating, and the right image shows a horizontal grating. Detailed Implementation

[0025] like Figure 1-3 As shown, a method for evaluating visual noise cancellation capability based on a contrast modulation grating includes the following steps:

[0026] S1. A grating noise image constructed based on a combination of grating size, grating contrast, grating spatial frequency, grating phase, number of pixels per viewing angle of the display, noise level, noise distribution, and noise contrast. The generation function of the grating target is:

[0027]

[0028] Where x, y are the coordinates of the image pixels, with the center of the image as the coordinate (0,0), C is the raster contrast ratio, S is the raster spatial frequency, P is the number of pixels per viewing angle of the display, θ is the raster angle, and γ is the raster phase.

[0029] In this embodiment, the grating size is 600x600 pixels, i.e., W = 600, and the grating angle is 90 degrees. Spatial frequency S = 0.5 cpd, number of pixels per viewing angle of the display is P = 64, raster phase γ = 0, initial raster contrast ratio C = 0.8, and noise contrast ratio is C. noise =0.2, noise distribution probability P n =0.4, and the contrast reduction factor is R = 1.0839. From the above parameters, we can obtain L max =32.

[0030] S2. Obtain a contrast-modulated grating image with specific noise, wherein the noise is represented in the form of square noise dots, and the noise is represented by three factors: the noise pixel size L, and the probability P. n Randomly distributed noise contrast correction value C noise ;

[0031] S21. Calculate the size of the noise pixels; the specific steps are as follows:

[0032] SA211. Calculate the maximum noise pixel width L that matches the grating spatial frequency S. max and round down:

[0033]

[0034] SA212, with L max Using 32 as the initial value, we find the first integer that is divisible by the image pixel size W by decreasing by -1. This integer is the noise pixel size L. In this scheme, we find the integer that is divisible by W = 600 by decreasing by 32, thus obtaining the noise pixel size L = 30.

[0035] S22. Divide the raster image into a grid; divide the raster image into equal horizontal and vertical segments, using noise pixel size L as the unit; for example... Figure 1 As shown.

[0036] S23. Generate noise points according to the noise distribution probability; for each of the above segmented regions, with probability P n The area is determined to be dark; otherwise, it is considered bright. If it is bright, the brightness of the area is increased by C. noise If it is dark, the brightness of that area decreases by C. noise .like Figure 2 As shown, in this scheme, P n The probability of 0.4 determines the brightness distribution of each noise point.

[0037] S3. The display screen showing the raster is facing the subject at a set distance, and the display brightness is set to a gray background at the maximum brightness of the display screen. Start the evaluation.

[0038] S4. After the image appears, the grating is randomly oriented. In this embodiment, for example... Figure 3 As shown, the image is displayed horizontally or vertically on the screen for 150 milliseconds before disappearing. The participant then answers the orientation of the grating. When the participant's answer matches the grating orientation, the contrast of the grating in the next test is [value missing]. C represents the raster contrast of the previous raster image, and R is a number greater than 1. The contrast is decreased until the subject cannot judge correctly. This contrast value is recorded as the subject's visual noise rejection ability score at that spatial frequency and noise contrast. In this scheme, the raster contrast C of the last test is the subject's visual noise rejection ability score at a spatial frequency of 0.5 cpd and a noise contrast of 0.2.

[0039] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for evaluating visual noise reduction capability based on a contrast modulation grating, characterized in that, Includes the following steps: S1. A grating noise image constructed based on a combination of grating size, grating contrast, grating spatial frequency, grating phase, number of pixels per viewing angle of the display, noise level, noise distribution, and noise contrast. The generation function for the grating target is: Where x, y are the coordinates of the image pixels, with the center of the image as the coordinate (0,0), C is the raster contrast ratio, S is the raster spatial frequency, P is the number of pixels per viewing angle of the display, θ is the raster angle, and γ is the raster phase. S2. Obtain a contrast-modulated grating image with specific noise, wherein the noise is represented in the form of square noise dots, and the noise is represented by three factors: the noise pixel size L, and the probability P. n Randomly distributed noise contrast correction value C noise ; S3. The display screen showing the raster is facing the subject at a set distance, and the display brightness is set to a gray background at the maximum brightness of the display screen. Start the evaluation. S4. After the image appears, the grating is randomly oriented and disappears from the screen after a set time. The subject then answers the orientation of the grating. When the subject's answer matches the grating orientation, the contrast of the grating in the next test is [value missing]. C is the raster contrast of the previous raster image, and R is a number greater than 1. The contrast is reduced until the subject cannot make a correct judgment. This contrast value is recorded as the subject's visual noise elimination ability score at this spatial frequency and noise contrast. The specific steps in step S2 to obtain a contrast-modulated grating image with specific noise are as follows: S21. Calculate the size L of the noise pixel; S22. Divide the raster image into a grid; divide the raster image into equal horizontal and vertical segments with noise pixel size L as the unit; S23. Generate noise points according to the noise distribution probability; for each of the above segmented regions, with probability P n The area is determined to be dark; otherwise, it is considered bright. If it is bright, the brightness of the area is increased by C. noise If it is dark, the brightness of that area decreases by C. noise; The specific steps for calculating the noise pixel size L are as follows: SA211, calculate the maximum matching noise pixel width L based on the grating spatial frequency S. max and round down: SA212, with L max Using the initial value as an integer, we decrease the value by -1 to find the first integer that can be divided by the image pixel size W, which is the noise pixel size L.

2. The method for evaluating visual noise reduction capability based on a contrast modulation grating according to claim 1, characterized in that, The orientation of the grating is random, either horizontal or vertical.

3. The method for evaluating visual noise reduction capability based on a contrast modulation grating according to claim 1, characterized in that, In step S4, the time is set to 150ms.

Images