Image color correction method for visual training

Abstract

The application is a kind of image color correction method for visual training, and relates to the technical field of visual training, which comprises the following steps: according to the principle of color brightness addition separation, a trainee wears color separation glasses, and under the condition of monocular observation, the critical hardware brightness level and the corresponding color that meet the image color separation condition when adjusting the image RGB channel are found respectively. In the application, the display device brightness, the color sensitivity of the trainee and the filtering condition of the color separation glasses for corresponding color light are tested and evaluated, and the color and brightness of the training image are adjusted accordingly, so as to eliminate the residual image phenomenon caused by incomplete color filtering of the color separation glasses and effectively improve the efficiency of binocular visual function training using the color separation glasses.

Description

Technical Field

[0001] This invention relates to the field of visual training technology, and more particularly to an image color correction method for visual training. Background Technology

[0002] Binocular vision training is an important part of visual training conducted in ophthalmology clinics, typically including simultaneous vision training, fusion vision training, and stereopsis training. During training, specific techniques are used to ensure that each of the trainee's eyes sees different images. The trainee is then required to adjust the position of the images or determine their attributes to complete a specific task. For example, in a simultaneous vision training task, the monitor displays both an image of a "birdcage" and an image of a "parrot." The trainee's left eye can only see the "birdcage" image, while their right eye can only see the "parrot" image. The trainee needs to drag the images to place the "parrot" inside the "birdcage."

[0003] For example, in a stereoscopic vision training task, two images containing text within boxes are displayed simultaneously on the monitor. The boxes in both images are the same size and completely overlap. The text in both images is also identical in size and content, but in one image, the text is positioned slightly to the left of the center of the box, while in the other image, the text is positioned slightly to the right of the center of the box. During training, the trainee's left eye can only see the image with the text slightly to the right of the center of the box, and the right eye can only see the image with the text slightly to the left of the center of the box. The trainee is required to merge the two slightly different images seen by each eye into one image and determine whether the text in the center of the box is "protruding" or "recessed" relative to the box.

[0004] A crucial prerequisite for the effective implementation of the above training is that the trainee's two eyes see different images separately. The trainee must process the different images seen by both eyes collaboratively to accurately complete the task. For example, in a simultaneous vision task, the trainee determines the true positions of two images separately and drags them to the appropriate location; in a stereoscopic vision task, the trainee determines the spatial depth relationship of the text relative to the box based on the parallax observed by both eyes (i.e., the degree of mutual deviation between the text positions in the two images).

[0005] A common method to make a trainee see different images for their left and right eyes is to use dichroic glasses (such as red-green, red-blue, or red-cyan glasses) and matching dichroic images. For example, the image of a "birdcage" is predominantly red, while the image of a "parrot" is predominantly blue-green. Theoretically, when the background is black, the trainee will only see the "birdcage" through the red lens of the dichroic glasses (usually the left eye), and only the "parrot" through the non-red lens (usually the right eye); when the background is white, the trainee will only see the "parrot" through the red lens of the dichroic glasses (usually the left eye), and only the "birdcage" through the non-red lens (usually the right eye).

[0006] However, in practical use, the method of separating images using color-separating glasses still has certain problems. Specifically, on a black background, when viewing an image on a screen using the red lens, one can see not only a predominantly red "birdcage" but also a faintly visible "parrot" with a predominantly blue-green hue; similarly, on a white background, when viewing an image on a screen using the red lens, one can see not only a predominantly blue-green "parrot" but also a faintly visible "birdcage" with a predominantly red hue. This incomplete image separation, or "ghosting," deviates from the established requirements of binocular vision training, severely interferes with training, and reduces the actual training effect.

[0007] Furthermore, in clinical binocular vision training, simple line drawings are often used initially to reduce difficulty and facilitate training for patients with impaired vision. However, in the later stages of training, complex training images that closely resemble real-life scenarios are needed to help patients apply their acquired binocular vision abilities to real-life situations. Therefore, the correction of complex, real-life scenario training images differs from that of simple line drawings. It cannot simply be a matter of adjusting the color brightness to a uniform value that does not produce color interference. Instead, it requires preserving the nuances of color content as much as possible without causing color interference, making it as close to real-life scenarios as possible.

[0008] To address the aforementioned issues, this proposal suggests an image color correction method for visual training. Summary of the Invention

[0009] The purpose of this invention is to provide an image color correction method for visual training in order to solve the above-mentioned problems.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: An image color correction method for visual training includes: S1. Obtain relevant information about the background color and color channels, including the RGB channel values ​​of the background color during training. , and And the maximum hardware brightness level that the RGB channels can use. , and And determine whether the most suitable color brightness separation property is "additive separation" or "subtractive separation"; S2. Based on the principle of additive color luminance separation, have the trainee wear color separation glasses and, under monocular observation conditions, find the critical hardware luminance level that satisfies the image color separation condition when adjusting any one of the RGB channels of the image individually. , and ; S3. Based on S2, further identify the critical hardware brightness level that meets the image color separation conditions when adjusting the GB channels of the image together. and ; S4. Based on the property of color brightness subtraction separation, have the trainee wear color separation glasses and, under monocular observation conditions, find the critical hardware brightness level that meets the image color display conditions when adjusting any one of the RGB channels of the image individually. , and ; S5. Based on S4, further identify the critical hardware brightness level that meets the image color display conditions when the image GB channels are jointly adjusted. and ; S6. Based on the most suitable color brightness separation property, use the measured parameters to correct the color brightness of the training image on the basis of the original training image.

[0011] Preferably, the RGB channel values ​​of the background color in S1 , and , which are the average values ​​of the RGB channels of the background image within the optimal placement area of ​​the training image.

[0012] Preferably, in S2, the determination is... , , and determined in S4 , and The method is as follows: Red Channel: Trainees wear color-separating glasses and use non-red lenses to observe a pattern on a display screen where only the red channel values ​​change continuously; When the red value of the pattern is higher than that of the background, find the critical point or range from distinguishable to indistinguishable, and record the minimum hardware brightness level at that point or range as _____. ; When the red value of the pattern is lower than that of the background, find the critical point or range, and record the maximum hardware brightness level at the critical point or range as [value missing]. ; Green Channel: Trainees wear color-separating glasses and use red lenses to observe patterns in which only the green channel values ​​change continuously; When the green value of the pattern is higher than that of the background, the minimum hardware brightness level within the critical point or range is recorded as: ; When the green value of the pattern is lower than that of the background, the maximum hardware brightness level within the critical point or range is recorded as: ; Blue Channel: Trainees wear color-separating glasses and use red lenses to observe patterns in which only the blue channel values ​​change continuously; When the blue value of the pattern is higher than that of the background, the minimum hardware brightness level within the critical point or range is recorded as: ; When the blue value of the pattern is lower than that of the background, the maximum hardware brightness level within the critical point or range is recorded as: .

[0013] Preferably, in S2, the determination is... , , The specific method is as follows: The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , [ ] is the endpoint, and the color blocks that change in descending order successively on the R channel are used by the trainee wearing color-separating glasses and using non-red lenses from [ ] , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the R channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , [ ] is the endpoint, and the color blocks that change in descending order successively on the G channel are used by the trainee wearing color-separating glasses and using the red lens to start from [ ] , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , The endpoint is a series of color blocks that change in descending order on channel B. Trainees are required to wear color-separating glasses and use the red lens to select the color blocks from […]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the B channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". .

[0014] Preferably, in S3, the determination is... and The specific method is as follows: Adjust the test background to: one line [ , , ], next line [ , , ], changing line by line; the background shows [ , , Starting from [], with [ , , [The target is] a series of color blocks that change continuously in descending order on the GB channel. The trainee is required to wear color-separating glasses and use the red lens to select from [...]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is The hardware brightness level corresponding to channel B That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". , .

[0015] Preferably, in S4, the determination is... , and The specific method is as follows: The test background is [ , , ], presented against a background of [0, , Starting from [], with [ , , [0, ] represents a series of color blocks that change in ascending order on the R channel. Trainees are required to wear color-separating glasses and use non-red lenses to start from [0, ] , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the R channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ ,0, Starting from [], with [ , , [ ] is the endpoint, and the color blocks change in ascending order successively on the G channel. Trainees are required to wear color-separating glasses and use the red lens to start from [ ] ,0, Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of that color block. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ;

[0016] The test background is [ , , ], presented on the background [ , Starting from [0], with [ , , [The target is a series of color blocks that change in ascending order on channel B, with the target being the end point. Trainees are required to wear color-separating glasses and use the red lens to select the color blocks from [The target is a series of color blocks that change in ascending order on channel B, with the target being the end point]. , Starting from [0], observe the color blocks one by one and determine the position where "the color block begins to be indistinguishable from the background". Then, determine the hardware brightness level corresponding to the B channel of that color block. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". .

[0017] Preferably, in S5, the determination is... and The specific method is as follows: Adjust the test background to: one line [ , , ], next line [ , , ], changing line by line; the background shows [ , , Starting from [], with [ , , [The target is] a series of color blocks that change continuously in ascending order on the GB channel. The trainee is required to wear color-separating glasses and use the red lens to select from [...]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is The hardware brightness level corresponding to channel B That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". , .

[0018] Preferably, the specific method for correcting the RGB channel values ​​of the training image in S6 is as follows: If the most suitable color-luminance separation property is additive separation, then the original R-channel value range of the training image visible to the red lens under the additive separation principle is linearly transformed to the range []. , The original GB channels remain unchanged; the original G channel value range of the training images visible to non-red lenses under the additive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; If the most suitable color-luminance separation property is subtractive separation, then the original G-channel value range of the training image visible to the red lens under the subtractive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; the original R channel value range of the training image (not visible to the red lens under the subtraction separation principle) is linearly transformed to the range []. , The original GB channel remains unchanged.

[0019] Preferably, during the test, the color block displayed on the screen is not smaller than a preset size, there is a color transition zone between the color block and the background, and the difference in hardware brightness level between adjacent pixels in the direction of color change within the transition zone does not exceed a preset value.

[0020] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. This invention evaluates the brightness of the display device, the color sensitivity of the trainee, and the filtering effect of the dichroic glasses on the corresponding colors of light through targeted testing. Based on this, the color and brightness of the training image are adjusted in a targeted manner to eliminate the afterimage phenomenon caused by the incomplete filtering of colors by the dichroic glasses, thereby effectively improving the efficiency of using dichroic glasses for binocular vision training.

[0021] 2. This invention can be used not only for training images with simple backgrounds, but also for complex training images with rich color levels and diverse content scenes. It can effectively preserve the complex background and content levels of training images, outlining object contours and related details, making the training scene closer to real life and improving training effectiveness. In cases of complex backgrounds, based on the average brightness of the background within the optimal placement area of ​​the training image, a more suitable color brightness separation property is dynamically selected, and the color and brightness of the training image are adjusted accordingly. This setting helps to select a more suitable color brightness variation space while ensuring effective separation of the training image, reflecting the layering of the training image content. Attached Figure Description

[0022] Further details, features, and advantages of this application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which: Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0023] Several embodiments of this application will now be described in more detail with reference to the accompanying drawings to enable those skilled in the art to implement this application. This application may be embodied in many different forms and for various purposes and should not be limited to the embodiments set forth herein. These embodiments are provided to make this application thorough and complete, and to fully convey the scope of this application to those skilled in the art. The embodiments described do not limit this application.

[0024] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It will be further understood that terms such as those defined in commonly used dictionaries shall be interpreted as having a meaning consistent with their meaning in the relevant field and / or the context of this specification, and shall not be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0025] Example 1

[0026] Its specific implementation method is combined with the appendix Figure 1 Please provide a detailed explanation.

[0027] In this embodiment, it includes: S1. Obtain relevant information about the background color and color channels, including the RGB channel values ​​of the background color during training. , and And the maximum hardware brightness level that the RGB channels can use. , and And determine whether the most suitable color brightness separation property is "additive separation" or "subtractive separation"; S2. Based on the principle of additive color luminance separation, have the trainee wear color separation glasses and, under monocular observation conditions, find the critical hardware luminance level that satisfies the image color separation condition when adjusting any one of the RGB channels of the image individually. , and ; S3. Based on S2, further identify the critical hardware brightness level that meets the image color separation conditions when adjusting the GB channels of the image together. and ; S4. Based on the property of color brightness subtraction separation, have the trainee wear color separation glasses and, under monocular observation conditions, find the critical hardware brightness level that meets the image color display conditions when adjusting any one of the RGB channels of the image individually. , and ; S5. Based on S4, further identify the critical hardware brightness level that meets the image color display conditions when the image GB channels are jointly adjusted. and ; S6. Based on the most suitable color brightness separation properties, use the parameters obtained from the above measurements to correct the color brightness of the training image on the basis of the original training image.

[0028] The above solution is applicable to scenarios where training images are displayed using electronic display devices and binocular vision training is conducted using color-separated glasses (e.g., red-green glasses, red-blue glasses, or red-cyan glasses). This solution is not applicable to situations where training images are not electronic (e.g., small devices used for binocular vision training called "red-green vector graphics"), or where training images are displayed electronically but image separation is not achieved using color-separated glasses (e.g., using a polarized light display to display images and observing them with polarized light glasses).

[0029] Furthermore, this solution is intended to eliminate "ghosting" in training images during training by using image color correction, based on an existing training scheme (including a defined display device, training images, and color-separating glasses), rather than for the complete design of new training images. Consequently, not all trainees need to perform color correction on their training images according to this solution; only those trainees whose original training scheme exhibits "ghosting" require this solution.

[0030] To precisely control the color and brightness of the training images, this solution first obtains information about the background color and the display device's color channels in S1. Since most displays currently use a combination of red, green, and blue to present color, the display device typically has three color channels: R, G, and B. For digitized background or training images, their color channels are consistent with those of the display device. However, for ease of image conversion, the image's R channel usually refers to a two-dimensional matrix, with the number of rows and columns corresponding to the image size. For example, if the image size is 600 pixels × 600 pixels, its R channel is a 600-row, 600-column matrix. The value of each element in the matrix represents the red brightness information contained in the pixel at the corresponding position in the image. It should be noted that the red brightness information represented by the elements in the matrix is ​​not the actual display brightness, but rather the hardware brightness level of the red channel that the pixel should use when displaying on the display device. For LED-backlit LCD displays, this hardware brightness level indicates the degree to which the filter valve at the red sub-pixel (containing a red filter) of that pixel is open when the display shows the image. The higher the degree of filter valve opening, the more backlight passes through the red filter and the filter valve. Correspondingly, more red light is emitted from that sub-pixel, making it appear brighter. The number of filter valve opening levels is usually determined by the bit depth of the display device. Therefore, the bit depth of the display device also determines the maximum hardware brightness level of the image color channels. When the display's bit depth is 8 bits, the total number of filter valve opening levels, or the total number of hardware brightness levels, is 2 to the power of 8, or 256; when the display's bit depth is 10 bits, the total number of filter valve opening levels, or the total number of hardware brightness levels, is 2 to the power of 10, or 1024. Since the starting value of a color channel is 0, when the display's bit depth is 8 bits, the maximum value of the color channel is... When the display has a bit depth of 10 bits, And so on.

[0031] After obtaining the R, G, and B channel matrices of the background image, the separation properties of the training image can be determined based on the actual situation of the RGB channels. When the background image color is simple, selection is relatively convenient. For example, when the background is white (i.e., the values ​​of the R, G, and B channels are all at their maximum values), only "subtractive separation" can be selected. That is, the color attributes of the training image (i.e., at least one of the R, G, and B channels has a value that has not reached its maximum value) will cause its brightness to be lower than the background. In this case, the brightness difference under dichroic glasses can be adjusted to make the training image either stand out (under dichroic glasses, the difference in brightness between the training image and the background is clearly visible; the greater the brightness difference, the clearer the image) or disappear into the background (under dichroic glasses, the difference in brightness between the training image and the background is very small and indistinguishable). Conversely, when the background is black, only "additive separation" can be selected. That is, the color attributes of the training image will cause its brightness to be higher than that of the background. In this case, you can increase the brightness of the training image to make it stand out from the background under dichroic glasses, or you can control the brightness of the training image to make it blend into the background under dichroic glasses and appear indistinguishable.

[0032] When the background image is complex (e.g., a gray background), or when one of the two training images is used as the background (e.g., training image A is "a large beach with a dog on the beach and a swimming ring on the ground next to it," and training image B is "a little girl," requiring B to be placed in the appropriate position of A to display "the little girl standing inside the swimming ring," in this task, training image A is essentially the background); or training image C is "a detached villa with gardens in front and behind, a bedroom on the second floor, and no railings on the bedroom window," and training image D is "railings on the second-floor bedroom window," requiring D to be placed in the appropriate position of C to display "railings installed on the outer edge of the second-floor bedroom window"), the determination of image separation properties becomes more complex. To obtain a wider color brightness variation space to preserve the brightness hierarchy of the image content, the average value of the background image in the RGB channels can be calculated. If this value exceeds 50% of the maximum hardware brightness level (e.g., when the display bit depth is 8 bits, if the background image R channel...), the determination becomes more complex. The mean is 130, G channel The mean is 190, B channel If the average value is 160, then the average value of the three channels is 160, which is higher than 50% of the maximum hardware brightness level value of 255. Therefore, "subtraction separation" is determined to be the optimal separation method for the training image, similar to the separation method for training images with a white background. If this value is lower than 50% of the maximum hardware brightness level value (for example, when the display bit depth is 8 bits, if the background image R channel...), then... The mean is 60, G channel The mean is 100, B channel If the average value of the three channels is 80 (which is 50% lower than the maximum hardware brightness level of 255), then "additive separation" is determined to be the optimal separation method for the training image, similar to the separation method for training images with a black background.

[0033] In addition, the above-mentioned "calculation based on the overall background image" , and The method described above is suitable for situations where the color brightness differences between different regions of the background are not significant (e.g., training image A mentioned above). If the color differences between different regions of the background are significant (e.g., training image C mentioned above), only the optimal placement area of ​​the training image (e.g., the "window of the second-floor bedroom" in training image C) can be selected to calculate the RGB channel values ​​of the background color. , and Preferably, this region is the optimal placement of the training image on the background, and the size of the region is the smaller of the two training images.

[0034] After determining the separation properties of the training images, it is necessary to further clarify the reasons for the "ghosting" that occurs during image separation in order to make targeted adjustments.

[0035] The phenomenon of image ghosting arises from a mismatch between the image display brightness, the dichroic glasses, and the sensitivity of the human eye. Theoretically, the red lens in dichroic glasses should completely filter out non-red light (including blue and green light), but in reality, this is not entirely true, and the filtering effect varies slightly depending on the lens material. Furthermore, in bright environments, the human eye is far more sensitive to green light than to red light (approximately 200 times); in dark environments, the human eye is far more sensitive to both green and blue light than to red light (nearly 250 times). This means that when an image with a predominantly blue-green hue is displayed on a black screen, even if 99.95% of the emitted blue-green light is filtered out by the red lens, the perceived brightness of the remaining blue-green light is still comparable to the perceived brightness of 10% red light. These perceived brightness effects from the residual blue-green light after passing through the red lens are the primary cause of "image ghosting" in actual training.

[0036] Therefore, the brightness of the displayed image (i.e., the actual brightness of the RGB colors in the image on the monitor), the dichroic lens (i.e., the type of material used and how much interfering light it filters out, usually measured as a percentage), and the sensitivity of the human eye (i.e., how much light is needed to produce a perceived brightness, and individual sensitivity varies) collectively determine the occurrence of "ghosting" after using dichroic glasses. When the image display brightness is high, or the dichroic lens's light filtering rate is low, or the human eye sensitivity is high, the "ghosting" phenomenon is more noticeable. When the image display brightness is low, or the dichroic lens's light filtering rate is high, or the human eye sensitivity is low, the "ghosting" phenomenon is less noticeable or even disappears. Since image display brightness, dichroic glasses, and human eye sensitivity all affect "ghosting," only by jointly correcting image display brightness, dichroic glasses, and human eye sensitivity to ensure that the residual light after interfering light passes through the dichroic lens is below the threshold that produces an additional brightness response can the "ghosting" phenomenon be effectively eliminated.

[0037] Based on the above analysis, after obtaining the basic information, a series of tests are needed to evaluate how the RGB channel values ​​of the training image should be adjusted on the current display device to avoid "ghosting." The trainee must participate in the evaluation, as "ghosting" is closely related to the trainee's sensitivity to red, green, and blue light. If the participant in the evaluation or testing is not the trainee, the evaluation or testing itself is invalid, and image color correction based on the evaluation or testing results cannot eliminate the trainee's "ghosting" phenomenon.

[0038] The following is a detailed analysis of the "ghosting" phenomenon, which serves as the main basis for determining the adjustment principles of RGB channel values ​​in training images.

[0039] Under the principle of "additive separation," when observing training images through a red lens, one can typically only observe training images that are "based on the background, with adjustments (additions) made only in the red channel," and not training images that are "based on the background, with adjustments (additions) made only in the blue-green channel." This is because, after being filtered by the red lens, most of the added red light is retained, while most of the added blue-green light is filtered out. Therefore, adjustments (additions) made only in the blue-green channel theoretically do not cause additional brightness perception when viewed through a red lens, and thus do not make the image stand out from the background. However, it is important to note that humans are extremely sensitive to green light. According to the visual function measured and recommended by the International Commission on Illumination (CIE), in daylight conditions, human sensitivity to green light is approximately 200 times that to red light, and in dark conditions, human sensitivity to both green and blue light is nearly 250 times that to red light. This means that if the adjustment (increase) made in the blue-green channel is too large, or even if the adjustment is small but the trainee's sensitivity to blue-green light is higher than that of the average person, then even if the red lens can filter out most of the blue-green light, the residual blue-green light may still cause additional brightness perception, thus making the image stand out from the background, manifested as the trainee seeing a "ghost image" under the red lens. Therefore, when using the "additive separation" method to separate training images, it is necessary to strictly control the upper limit of the GB channel value of "theoretically invisible images under the red lens," so that the blue-green light emitted by the display device, after being filtered by the color-separating glasses, still cannot reach the trainee's perception threshold (i.e., the trainee cannot distinguish the difference in brightness between the image and the background).

[0040] Under the "additive separation" principle, when observing training images through a non-red lens, one can typically only observe training images that are "adjusted (added) to the blue-green channel based on the background," and not training images that are "adjusted (added) only to the red channel based on the background." This is because non-red lenses filter out the vast majority of red light; therefore, when observed through a non-red lens, images that are adjusted (added) only to the red channel will theoretically be confused with the background and indistinguishable. Furthermore, since the human eye is not very sensitive to red light, the probability of ghosting when observing images that are "adjusted (added) only to the red channel" through a non-red lens is extremely small, far less than the aforementioned probability of "seeing ghosting in images that are 'adjusted (added) only to the blue-green channel' through a red lens." However, it should be noted that for a small number of individuals who are highly sensitive to red light, ghosting may still occur during training. In other words, when using the "additive separation" method to separate training images, it is also necessary to control the upper limit of the R channel value of "theoretically invisible images under non-red lenses" so that the red light emitted by the display device, after being filtered by the color separation glasses, still cannot reach the trainee's perception threshold (i.e., the trainee cannot distinguish the difference in brightness between the image and the background).

[0041] Under the "subtractive separation" principle, when observing through a red lens, one can typically only observe training images that are "based on the background, with adjustments (reductions) made only in the red channel," and not training images that are "based on the background, with adjustments (increases) made only in the blue-green channel." This is because training images that are "adjusted (reduced) only in the red channel" emit less red light compared to the background, and red light is less affected by the red lens's filtering. Therefore, the difference in the amount of red light emitted between the image and the background is largely retained after filtering, resulting in a noticeable and perceptible brightness difference between the image and the background (the background is brighter, the image is darker), making the image stand out from the background. On the other hand, compared to the background, training images that are "adjusted (reduced) only in the blue-green channel" emit the same amount of red light, only the blue-green light is reduced. Because blue-green light is significantly affected by red lenses, the difference in blue-green light emission between the image and the background is only partially retained (generally no more than 1%) after filtering by non-red lenses. Therefore, in most cases, this difference in blue-green light emission will not cause a noticeable, perceptible change in brightness, and thus will not cause the image to stand out from the background. However, as mentioned earlier, because the human eye is highly sensitive to blue-green light, if the adjustment (reduction) in the blue-green channel is too large, or even if the adjustment is small but the trainee's sensitivity to blue-green light is higher than average, then even if the red lens filters out the vast majority of blue-green light, the residual blue-green light may still cause additional brightness perception, causing the image to stand out from the background. This manifests as the trainee seeing a "ghost image" of the image under the red lens. Therefore, it is necessary to strictly control the lower limit of the GB channel value, which theoretically makes the image invisible under red lenses, ensuring it does not fall below a specific critical value.

[0042] Under the "subtractive separation" principle, when observing images through a non-red lens, one can typically only observe training images that are "based on the background, with adjustments (increases) only in the blue-green channels," and not training images that are "based on the background, with adjustments (decreases) only in the red channel." This is because training images that are "adjusted (decreased) only in the blue-green channels" emit less blue-green light compared to the background. Since blue-green light is less affected by the filtering of non-red lenses, the difference in the amount of blue-green light emitted between the image and the background is largely retained after filtering. This results in a noticeable and perceptible brightness difference between the red image and the background (the background is brighter, the red image is darker), making the image stand out from the background. On the other hand, compared to the background, training images that are "adjusted (decreased) only in the blue-green channels" emit the same amount of blue-green light, only the red light is reduced. Because red light is significantly affected by non-red lenses, the difference in red light emission between the image and the background is only partially retained after filtering by the non-red lenses. Furthermore, due to the low sensitivity of the human eye to red light, this difference in red light emission between the image and the background hardly causes a noticeable change in brightness, making it difficult for the image to stand out from the background. However, it's important to note that for a few individuals who are highly sensitive to red light, image retention may still occur during training. In other words, when using the "subtractive separation" method to separate training images, it's crucial to control the lower limit of the R-channel value for images that are theoretically invisible under non-red lenses, ensuring it doesn't fall below a specific critical value.

[0043] Based on the above analysis, it can be understood that S2 determines , , and determined in S4 , and The method should be: Trainees wear color-separating glasses and observe a pattern on a display screen—based on the background and continuously varying only in the red channel—using non-red lenses. When the pattern's red channel value is higher than the background's, the critical point or range where the pattern and background become indistinguishable is identified. The hardware brightness level corresponding to the red channel at the critical point, or the minimum hardware brightness level corresponding to the red channel within the critical range, is determined as [the threshold value]. When the red channel value of the pattern is lower than that of the red channel value of the background, find the critical point or range at which the pattern and background become indistinguishable, and determine the hardware brightness level corresponding to the red channel at the critical point or the maximum hardware brightness level corresponding to the red channel within the critical range. ; Trainees wear color-separation glasses and observe a pattern on a display screen—based on the background and continuously varying only in the green channel—using red lenses. When the green channel value of the pattern is higher than that of the background, the critical point or range where the pattern and background become indistinguishable is identified. The hardware brightness level corresponding to the green channel at the critical point, or the minimum hardware brightness level corresponding to the green channel within the critical range, is determined as [the threshold value]. When the green channel value of the pattern is lower than that of the green channel value of the background, find the critical point or range at which the pattern and background become indistinguishable, and determine the hardware brightness level corresponding to the green channel at the critical point or the maximum hardware brightness level corresponding to the green channel within the critical range. ; Trainees wear color-separation glasses and observe a pattern on a display screen—based on the background and continuously varying only in the blue channel—using red lenses. When the blue channel value of the pattern is higher than that of the background, the critical point or range where the pattern and background become indistinguishable is identified. The hardware brightness level corresponding to the blue channel at the critical point, or the minimum hardware brightness level corresponding to the blue channel within the critical range, is determined as [the threshold value]. When the blue channel value of the pattern is lower than that of the blue channel value of the background, find the critical point or range at which the pattern and background become indistinguishable, and determine the hardware brightness level corresponding to the blue channel at the critical point or the maximum hardware brightness level corresponding to the blue channel within the critical range. .

[0044] Theoretically, there are many methods to find the critical point where the pattern and background become indistinguishable. This solution provides a method that is convenient for trainees and can be completed quickly and easily, specifically as follows: The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , [ ] is the endpoint, and the color blocks that change in descending order successively on the R channel are required for the trainee to wear color-separating glasses and use non-red lenses to start from [ ] , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the R channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , [ ] is the endpoint, and the color blocks change in descending order successively on the G channel. Trainees are required to wear color-separating glasses and use the red lens to select from [ ] , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , The endpoint is a series of color blocks that change in descending order on channel B. Trainees are required to wear color-separating glasses and use the red lens to select the color blocks from […]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the B channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against a background of [0, , Starting from [], with [ , , [0, ] represents a series of color blocks that change in ascending order on the R channel. Trainees are required to wear color-separating glasses and use non-red lenses to start from [0, ] , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the R channel of the color block at that position. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ ,0, Starting from [], with [ , , [ ] is the endpoint, and the color blocks change in ascending order successively on the G channel. Trainees are required to wear color-separating glasses and use the red lens to start from [ ] ,0, Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of that color block. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented on the background [ , Starting from [0], with [ , , [The target is a series of color blocks that change in ascending order on channel B, with the target being the end point. Trainees are required to wear color-separating glasses and use the red lens to select the color blocks from [the target]. , Starting from [0], observe the color blocks one by one and determine the position where "the color block begins to be indistinguishable from the background". Then, determine the hardware brightness level corresponding to the B channel of that color block. That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". .

[0045] Based on the above analysis and testing methods, it can be seen that the critical value The upper limit (maximum value) of the R channel of a training image that "changes only in the blue-green channel based on the background, is visible under non-red lenses, and invisible under red lenses," is used to ensure that the image does not exhibit ghosting under the condition of "additive separation + red lens observation"; the obtained critical value The lower bound (minimum value) of the R channel of a training image is used to determine whether it "changes only in the blue-green channel based on the background, is visible under non-red lenses, and is invisible under red lenses," to ensure that the image does not exhibit ghosting under the condition of "subtraction separation + red lens observation." Based on this, and The range between these values ​​can be defined as the selectable range of the R channel values ​​for this image. Clearly, the selectable range of R channel values ​​provides feasibility for setting different content and multiple layers of red within the image.

[0046] Similarly, the critical values ​​obtained during the above testing process and These threshold values ​​are used to determine the upper limits (maximum values) of the G and B channels of training images that "change only in the red channel based on the background, are visible under red lenses, and are invisible under non-red lenses," to ensure that the image does not exhibit ghosting under conditions of "additive separation + observation under non-red lenses + use of either the G or B channel alone." and These are used to determine the lower limits (minimum values) of the G and B channels for training images that "change only in the red channel based on the background, are visible under red lenses, and are not visible under non-red lenses," to ensure that the image will not exhibit ghosting under the conditions of "subtraction separation + observation under non-red lenses + use of either the G or B channel alone." Based on this, and The range between these values ​​can be defined as the selectable range of G channel values ​​when the image uses the G channel alone (i.e., when all B channel values ​​are set to 0). and The range can be defined as the selectable range of B channel values ​​when the image uses only the B channel (i.e., when all G channel values ​​are set to 0).

[0047] Similarly, the selectable range of G and B channel values ​​provides feasibility for setting multiple layers of different content and blue-green hues within the image.

[0048] It should be noted that the "descending order change" or "ascending order change" in the above testing method cannot be chosen arbitrarily. This is because the testing method has a certain impact on the test results. Generally, due to the existence of inertia, the test results in the descending order change sequence will be smaller. That is, since the background and image are "distinguished" at the beginning of the test, in subsequent judgments, they will be habitually judged as "distinguished," leading to most color blocks between "distinguished" and "indistinguishable" being determined as "distinguished," until the difference between the color block and the background is reduced to a very small degree before changing to "indistinguishable." Similarly, the test results in the ascending order change sequence will be larger. Therefore, based on the principle of ghosting, using ascending order change in the test to determine the lower limit (minimum value) of the channel value, and using descending order change in the test to determine the upper limit (maximum value) of the channel value, is a more reliable choice and more conducive to eliminating ghosting. In addition, if multiple tests are conducted, the critical value may fluctuate slightly, which is a normal phenomenon because human physiological state also fluctuates. To ensure the elimination of ghosting, it is recommended to selectively use the minimum or maximum value of the critical range, based on the principle behind ghosting. For example, when determining... , and When selecting the minimum value (descending order change), in order of determination , and The maximum value is selected (in ascending order).

[0049] In descending or ascending order changes, the spacing between channel values ​​of color blocks does not need to be very strict; it just needs to be as uniform as possible. For example, when the R channel value changes continuously in descending order, it can be [ , , ]、[ , , ]、[ , , ..., or it could be [ , , ]、[ , , ]、[ , , ..., n is a positive integer, and the value can be adjusted according to the actual situation.

[0050] In addition, to ensure that trainees can accurately perceive the color and brightness of the color blocks and avoid the impact of boundary effects (i.e., the difference in color brightness appears larger on both sides of a boundary where color brightness changes abruptly; this phenomenon of "the difference appearing larger" does not occur during gradual changes in color brightness) on the test, the size of the color blocks displayed on the screen during the test is no smaller than a preset size, and there is a color transition zone between the color blocks and the background. Within the transition zone, the difference in hardware brightness levels between adjacent pixels in the direction of color change does not exceed a preset value. Preferably, the preset size is 2 degrees of viewing angle × 2 degrees of viewing angle, and the preset value is 3.

[0051] It should be noted that if both training images are simply "adjusted based on the background, adjusting one of the three RGB color channels", then the above... , and The test results are sufficient to meet the correction requirements for the training images. However, in most cases, the GB channels in the training images will be adjusted simultaneously. In this case, further adjustments are needed. and Based on this, the critical hardware brightness level that meets the image color separation conditions when the GB two channels are adjusted together was found. and The specific method is as follows: Adjust the test background to: one line [ , , ], next line [ , , ], changing line by line; the background shows [ , , Starting from [], with [ , , [The target is] a series of color blocks that change continuously in descending order on the GB channel. The trainee is required to wear color-separating glasses and use the red lens to select from [...]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is The hardware brightness level corresponding to channel B That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". , .

[0052] Similarly, it is possible to and Based on this, the critical hardware brightness level that meets the image color separation conditions when the GB two channels are adjusted together was found. and The specific method is as follows: Adjust the test background to: one line [ , , ], next line [ , , ], changing line by line; the background shows [ , , Starting from [], with [ , , [The target is] a series of color blocks that change continuously in ascending order on the GB channel. The trainee is required to wear color-separating glasses and use the red lens to select from [...]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is The hardware brightness level corresponding to channel B That is If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". , .

[0053] It's important to note that human brightness perception exhibits temporal and spatial averaging characteristics. For example, when a light flickers slowly, we can clearly perceive that "the light was on one second and went off the next." However, when the light flickers very quickly (e.g., 100 times per second), our visual perception cannot keep up, and we only perceive that "the light is always on, just less bright than usual." This is an example of brightness averaging over time. Similarly, if some areas on a large screen are damaged and not lighting up, these damaged pixels can be seen up close, but from a distance, they become unclear. We only perceive that "the screen is still bright, just a little less bright than usual." This is an example of brightness averaging over space.

[0054] Therefore, the test background was adjusted to "one line [ , , ], next line [ , , "Change line by line" means adjusting the brightness of the entire background to [ ]. , , ] Corresponding brightness and [ , , The average value corresponding to the brightness.

[0055] Based on this, the test results were obtained and This can be used to set the maximum values ​​of the G and B channels when jointly adjusting the GB channels of the training image on top of the background under additive separation conditions, to ensure that no ghosting of the training image appears when viewed with a red lens. Similarly, the test results obtained... and It can be used to set the minimum values ​​of the G and B channels in the training image under subtraction separation, so as to ensure that no afterimage of the training image will appear when viewed with a red lens.

[0056] In the above tests, the starting point can be one of the previously mentioned […]. , , ]and[ , , ], or it can be [ , , ]and[ [0, 0], each has its own advantages. For example, choosing [ , , When used as a starting point for testing, due to the difference between this starting point and the background, it is more effective than [[]. , , The difference between the color block and the background is greater, so test takers have a clearer feeling about the state where "the color block is distinguishable from the background." Therefore, they are more confident in confirming that "the color block is initially indistinguishable from the background," and are less likely to make a selection bias due to hesitation. However, because [ , , The distance between the test starting point and the test result is [ ] compared to [ , , The distance between the color block and the test result is greater, so it takes more time to select the color block and the testing efficiency is relatively lower.

[0057] Furthermore, in descending or ascending order adjustments of the GB two channels, the two channels can be adjusted simultaneously or alternately. However, when adjusting alternately, it is recommended to control the total adjustment value of a single channel within one alternation cycle to ensure the most accurate measurement of the critical value or critical range. Of course, channel values ​​exceeding the critical range that meet the color separation conditions can also be used. For example, A value smaller than the actual critical range can be used. A value larger than the actual critical range can be used. However, using a value outside the critical range will lead to […]. , The reduced range of the color space affects the correction of color content levels within the G channel. When performing color level correction on the original training image, [ , The larger the range of the G channel, the greater the space for adjusting the color content levels within the G channel, and the more completely the color content levels are preserved after correction. Similarly, the same applies to the R and B channels.

[0058] After all tests are completed, the RGB channel values ​​of the original training images can be corrected based on the various threshold values ​​obtained from the tests. The specific method is as follows: If the most suitable color-luminance separation property is additive separation, then the original R-channel value range of the training image visible to the red lens under the additive separation principle is linearly transformed to the range []. , The original GB channels remain unchanged; the original G channel value range of the training images visible to non-red lenses under the additive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; If the most suitable color-luminance separation property is subtractive separation, then the original G-channel value range of the training image visible to the red lens under the subtractive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; the original R channel value range of the training image (not visible to the red lens under the subtraction separation principle) is linearly transformed to the range []. , The original GB channel remains unchanged.

[0059] It should be noted that, in order to better reflect the gradation of color content, the method for correcting the RGB channel values ​​of the training image can also be: If the most suitable color-luminance separation property is additive separation, then the original R-channel value range of the training image visible to the red lens under the additive separation principle is linearly transformed to the range []. , The original GB channels remain unchanged; the original G channel value range of the training images visible to non-red lenses under the additive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; If the most suitable color-luminance separation property is subtractive separation, then the original G-channel value range of the training image visible to the red lens under the subtractive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; the original R channel value range of the training image (not visible to the red lens under the subtraction separation principle) is linearly transformed to the range []. , The original GB channel remains unchanged.

[0060] The specific formula for linear change is: ; Where V represents the original RGB channel values ​​of the image, V' represents the RGB channel values ​​after linear transformation, Vmin represents the minimum value among the original RGB channel values, Vmax represents the maximum value among the original RGB channel values, and V'min represents the minimum value among the RGB channel values ​​after linear transformation (for example, ...). V'max is the maximum value of the RGB channel values ​​of the image after the linear transformation (for example, ...). Essentially, the above conversion is to proportionally convert Vmin~Vmax into V'min~V'max.

[0061] Additionally, during training, the trainer needs to drag the images to make the two images overlap in space. The specific method for setting the RGB channel values ​​of the training images when overlapping is as follows: If the training content consists of simple images that do not require a sense of color hierarchy (images often used in the early stages of training, such as line drawings), then in the case of additive separation, the RGB channel values ​​of the overlapping part of the image are the larger values ​​of the R channel, G channel, and B channel of the overlapping part of the two training images, respectively; in the case of subtractive separation, the RGB channel values ​​of the overlapping part of the image are the smaller values ​​of the R channel, G channel, and B channel of the overlapping part of the two training images, respectively.

[0062] If the training content consists of complex images that require a sense of color depth (images frequently used in the later stages of training, such as landscape and portrait photos), due to the significant differences in color depth between different training images, the only option is to use an overlay principle to set the RGB channel values ​​of the overlapping areas. That is, the image being dragged is set as the "upper layer image," and only the RGB channel values ​​of the corresponding area of ​​the "upper layer image" are used to display the content within that area of ​​the overlapping region.

[0063] Finally, the testing and image setting adjustment methods described above are applicable to the vast majority of color separation glasses currently on the market. Whether it's red-green, red-blue, or red-cyan glasses, the above methods can ensure the effectiveness of "eliminating afterimages and achieving perfect image separation."

[0064] In summary, this solution not only effectively eliminates image ghosting during training, but also preserves more color and brightness levels in the original training images, better outlining object contours and related details, making training more relevant to real life and achieving better training results.

[0065] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0066] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely for distinguishing one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0067] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0068] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0069] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An image color correction method for visual training, characterized in that, include: S1. Obtain relevant information about the background color and color channels, including the RGB channel values ​​of the background color during training. , and And the maximum hardware brightness level that can be used for the RGB channels. , and And determine whether the most suitable color brightness separation property is "additive separation" or "subtractive separation"; S2. Based on the principle of additive color-brightness separation, trainees wear color separation glasses and, under monocular observation conditions, find the critical hardware brightness level that satisfies the image color separation condition when adjusting any one of the RGB channels of the image individually. , and ; S3. Based on S2, further identify the critical hardware brightness level that meets the image color separation conditions when adjusting the GB channels of the image together. and ; S4. Based on the property of color brightness subtraction separation, have the trainee wear color separation glasses and, under monocular observation conditions, find the critical hardware brightness level that meets the image color display conditions when adjusting any one of the RGB channels of the image individually. , and ; S5. Based on S4, further identify the critical hardware brightness level that meets the image color display conditions when adjusting the image's GB channels in combination. and ; S6. Based on the most suitable color brightness separation property, use the measured parameters to correct the color brightness of the training image on the basis of the original training image.

2. The image color correction method for visual training according to claim 1, characterized in that, RGB channel values ​​of the background color in S1 , and , which are the average values ​​of the RGB channels of the background image within the optimal placement area of ​​the training image.

3. The image color correction method for visual training according to claim 1, characterized in that, Determined in S2 , , and determined in S4 , and The method is as follows: Red Channel: Trainees wear color-separating glasses and use non-red lenses to observe a pattern on a display screen where only the red channel values ​​change continuously; When the red value of the pattern is higher than that of the background, find the critical point or range from distinguishable to indistinguishable, and record the minimum hardware brightness level at that point or range as _____. ; When the red value of the pattern is lower than that of the background, find the critical point or range, and record the maximum hardware brightness level at the critical point or range as [value missing]. ; Green Channel: Trainees wear color-separating glasses and use red lenses to observe patterns in which only the green channel values ​​change continuously; When the green value of the pattern is higher than that of the background, the minimum hardware brightness level within the critical point or range is recorded as: ; When the green value of the pattern is lower than that of the background, the maximum hardware brightness level within the critical point or range is recorded as: ; Blue Channel: Trainees wear color-separating glasses and use red lenses to observe patterns in which only the blue channel values ​​change continuously; When the blue value of the pattern is higher than that of the background, the minimum hardware brightness level within the critical point or range is recorded as: ; When the blue value of the pattern is lower than that of the background, the maximum hardware brightness level within the critical point or range is recorded as: .

4. The image color correction method for visual training according to claim 3, characterized in that, Determined in S2 , , The specific method is as follows: The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , [ ] is the endpoint, and the color blocks that change in descending order successively on the R channel are used by the trainee wearing color-separating glasses and using non-red lenses from [ ] , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the R channel of the color block at that position. That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , [ ] is the endpoint, and the color blocks that change in descending order successively on the G channel are used by the trainee wearing color-separating glasses and using the red lens to start from [ ] , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ , , Starting from [], with [ , , The endpoint is a series of color blocks that change in descending order on channel B. Trainees are required to wear color-separating glasses and use the red lens to select the color blocks from […]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the B channel of the color block at that position. That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". .

5. The image color correction method for visual training according to claim 4, characterized in that, Determined in S3 and The specific method is as follows: Adjust the test background to: one line [ , , ], next line [ , , ], changing line by line; the background shows [ , , Starting from [], with [ , , [ ] is the endpoint, and the trainee is required to wear color-separating glasses and use the red lens to select from [ ] a series of color blocks that change continuously in descending order on the GB channel. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is The hardware brightness level corresponding to channel B That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". , .

6. The image color correction method for visual training according to claim 5, characterized in that, Determined in S4 , and The specific method is as follows: The test background is [ , , ], presented against a background of [0, , Starting from [], with [ , , [0, ] represents a series of color blocks that change in ascending order on the R channel. Trainees are required to wear color-separating glasses and use non-red lenses to start from [0, ] , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the R channel of the color block at that position. That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented against the background with [ ,0, Starting from [], with [ , , [ ] is the endpoint, and the color blocks change in ascending order successively on the G channel. Trainees are required to wear color-separating glasses and use the red lens to start from [ ] ,0, Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of that color block. That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". ; The test background is [ , , ], presented on the background [ , Starting from [0], with [ , , [The target is a series of color blocks that change in ascending order on channel B, with the target being the end point. Trainees are required to wear color-separating glasses and use the red lens to select the color blocks from [the target]. , Starting from [0], observe the color blocks one by one and determine the position where "the color block begins to be indistinguishable from the background". Then, determine the hardware brightness level corresponding to the B channel of that color block. That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". .

7. The image color correction method for visual training according to claim 6, characterized in that, S5 determines and The specific method is as follows: Adjust the test background to: one line [ , , ], next line [ , , ], changing line by line; the background shows [ , , Starting from [], with [ , , [The target is] a series of color blocks that change continuously in ascending order on the GB channel. The trainee is required to wear color-separating glasses and use the red lens to select from [...]. , , Begin by observing the color blocks one by one, and determine the position where the color block becomes indistinguishable from the background. Then, determine the hardware brightness level corresponding to the G channel of the color block at that position. That is The hardware brightness level corresponding to channel B That is ; If the endpoint color block [ , , It also fails to meet the condition that "the color block and the background are indistinguishable". , .

8. The image color correction method for visual training according to claim 7, characterized in that, The specific method for correcting the RGB channel values ​​of training images in S6 is as follows: If the most suitable color-luminance separation property is additive separation, then the original R-channel value range of the training image visible to the red lens under the additive separation principle is linearly transformed to the range []. , The original GB channels remain unchanged; the original G channel value range of the training images visible to non-red lenses under the additive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; If the most suitable color brightness separation property is subtractive separation, then the original G channel value range of the training image visible to the red lens under the subtractive separation principle is linearly transformed to the range []. , The original B channel numerical range is linearly transformed to the range [ , The original R channel remains unchanged; the original R channel value range of the training image (not visible to the red lens under the subtraction separation principle) is linearly transformed to the range []. , The original GB channel remains unchanged.

9. The image color correction method for visual training according to claim 8, characterized in that, During testing, the color blocks displayed on the screen are no smaller than the preset size, there is a color transition zone between the color blocks and the background, and the difference in hardware brightness level between adjacent pixels in the direction of color change within the transition zone does not exceed the preset value.

Images