Amblyopia training system and control method thereof

Abstract

The invention provides an amblyopia training system and a control method thereof, and the amblyopia training system comprises a camera acquisition module, a calibration module, an amblyopia training module, a hyperopia defocus parameter modeling module, a calculation unit and a display unit. The system obtains pupil data of a user through the camera acquisition module, the calibration module calculates and displays calibration parameters, and the amblyopia training module generates basic training images and training tasks. And the hyperopia defocusing parameter modeling module performs defocusing processing on the training image based on the calibration parameter and the pupil data to generate a composite training image with a clear center and a peripheral hyperopia defocusing effect. The calculation unit performs real-time rendering, and the display unit displays the composite training image according to the training task. According to the invention, the cooperation of visual training and eye axis regulation and control is realized, and an effective comprehensive treatment scheme is provided for hyperopia amblyopia patients.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to a weak vision training system and a control method thereof. BACKGROUND

[0002] As a common developmental eye disease in children, the traditional treatment methods of amblyopia mainly include refractive correction, occlusion therapy and fine eye training. Although these methods can improve vision to some extent, they have the problems of long treatment cycle and limited effect, especially for hypermetropic amblyopia patients, the problem of delayed orthoptics process caused by delayed eye axis development has not been effectively solved.

[0003] The existing digital amblyopia training technology mainly presents high-contrast graphics or dynamic visual targets to improve the central visual acuity of the amblyopic eye, but these technologies cannot distinguish the different needs of the central and peripheral regions, and cannot simultaneously improve vision and regulate eye axis development, resulting in unsatisfactory treatment effect and failing to meet the special needs of hypermetropic amblyopia patients to accelerate eye axis development. SUMMARY

[0004] The main purpose of the present application is to solve the technical problem that the existing amblyopia training technology cannot simultaneously realize vision improvement and eye axis regulation, especially cannot provide effective means to promote the development of hypermetropic amblyopia patients; The present application provides an amblyopia training system, which comprises: A camera acquisition module for acquiring pupil data of a user; A calibration module for calculating display calibration parameters according to hardware parameters of a display unit; An amblyopia training module for generating a basic training image and a training task; A hypermetropic defocus parameter modeling module for receiving the display calibration parameters and the pupil data, performing defocus processing on the basic training image, and generating a composite training image; A computing unit for performing real-time rendering processing on the composite training image and outputting an image display signal; A display unit for receiving the image display signal and the training task, and displaying the composite training image according to the training task to realize the synchronous performance of central clear vision training and peripheral hypermetropic defocus stimulation.

[0005] The present application also provides a control method of an amblyopia training system, which comprises: Receiving hardware parameters of a display unit through the calibration module, calculating and outputting display calibration parameters; Acquiring pupil data of a user through the camera acquisition module; The weak vision training module generates a basic training image and a training task of a corresponding stage according to a preset training process. The hyperopic defocus parameter modeling module receives the display calibration parameter and the pupil data, calculates an image blur parameter in a defocus state, and performs defocus processing on the basic training image based on the image blur parameter to generate a composite training image. The computing unit performs real-time rendering processing on the composite training image and outputs an image display signal. The display unit receives the image display signal and the training task, displays the composite training image according to the training task, and realizes the synchronous performance of central clear vision training and peripheral hyperopic defocus stimulation. The weak vision training module monitors the training progress in real time, records the training data, and dynamically adjusts the training parameters and process according to the training effect.

[0006] The weak vision training system and the control method thereof, the weak vision training system comprises a camera acquisition module, a calibration module, a weak vision training module, a hyperopic defocus parameter modeling module, a computing unit and a display unit. The system obtains user pupil data through the camera acquisition module, calculates display calibration parameters through the calibration module, and generates basic training images and training tasks through the weak vision training module. The hyperopic defocus parameter modeling module performs defocus processing on the training image based on the calibration parameter and the pupil data to generate a composite training image with central clear and peripheral hyperopic defocus effect. The computing unit performs real-time rendering, and the display unit displays the composite training image according to the training task. The present application realizes the cooperation of vision training and eye axis regulation, and provides an effective comprehensive treatment scheme for hyperopic amblyopia patients.

[0007] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application will be realized and achieved by the structure particularly pointed out in the description, claims and drawings.

[0008] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the following preferred embodiments are specifically described below, and the accompanying drawings are described in detail as follows. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 An embodiment schematic diagram of a weak vision training system in the embodiments of the present application; Figure 2 An embodiment schematic diagram of a weak vision training method in the embodiments of the present application. DETAILED DESCRIPTION

[0010] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the present application will be described clearly and completely below with reference to the drawings, obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative work belong to the protection scope of the present application.

[0011] The terms "comprising" and "having" and any variations thereof mentioned in the embodiments of the present application are intended to cover the inclusions not exclusively. For example, the process, method, system, product or device end comprising a series of steps or units is not limited to the listed steps or units, but optionally also includes other steps or units not listed, or optionally also includes other steps or units inherent to the process, method, product or device end.

[0012] In order to facilitate the understanding of the embodiments, first, a weak vision training system disclosed by the embodiments of the present application is introduced in detail. The weak vision training system comprises: The camera acquisition module 101 is configured to acquire pupil data of a user. The calibration module 102 is configured to calculate display calibration parameters according to hardware parameters of a display unit. The amblyopia training module 103 is configured to generate a basic training image and a training task. The hyper-focal parameter modeling module 104 is configured to receive the display calibration parameters and the pupil data, perform defocus processing on the basic training image, and generate a composite training image. The computing unit 105 is configured to perform real-time rendering processing on the composite training image, and output an image display signal. The display unit 106 is configured to receive the image display signal and the training task, and display the composite training image according to the training task, so as to realize the synchronous performance of central clear vision training and peripheral hyper-focal stimulation.

[0013] In the embodiments, the hardware parameters include screen width, screen height, observation distance, resolution and front lens parameters, and the display calibration parameters include pixel number per degree of viewing angle, pixel density, field of view angle and physical pixel conversion factor.

[0014] Specifically, the screen width and screen height in the hardware parameters refer to the physical dimensions of the display unit 106, usually in millimeters or centimeters, which directly determine the size of the visible area and provide basic data for subsequent field of view angle calculation. The observation distance refers to the vertical distance from the user's eyes to the display screen, which plays a key role in determining the actual viewing angle range of the user. Different observation distances will directly affect the user's perception of the image. The resolution parameter represents the pixel density of the display screen, including the number of pixels in the horizontal and vertical directions. This parameter is used in combination with the physical dimensions of the screen to calculate the pixel distribution density per unit length. The front lens parameter refers to the optical characteristics such as diopter and focal length of the optical lens that may exist in front of the display unit 106. When the system is equipped with a front lens, the display effect needs to be corrected accordingly.

[0015] The display calibration parameters are key technical parameters calculated based on the hardware parameters. The pixels per degree PPD represents the number of pixels corresponding to each degree of viewing angle, which establishes a corresponding relationship between angular vision and pixel resolution and is a core parameter for implementing precise defocus control. The pixel density PPI represents the number of pixels per inch, reflecting the fineness of the display unit 106. The field of view angle FOV includes the horizontal and vertical field of view angles, representing the maximum viewing angle range that the user can observe at the current observation distance, which directly affects the division and processing range of the defocus area. The physical pixel conversion factor is used to accurately convert between physical dimensions and pixel coordinates and is a bridge parameter connecting the results of geometric optics calculation and digital image processing.

[0016] In this embodiment, the calibration module 102 is specifically configured to: According to the screen width, screen height, and resolution, calculate the pixel density in the horizontal and vertical directions; According to the observation distance and pixel density, calculate the pixels per degree of viewing angle; According to the screen width, screen height, and observation distance, calculate the horizontal and vertical field of view angles through geometric relationships; According to the screen width, screen height, and resolution, calculate the conversion factor between physical dimensions and pixel coordinates as the physical pixel conversion factor; When the front lens parameter exists, perform optical correction on the pixels per degree of viewing angle, pixel density, field of view angle, and physical pixel conversion factor results according to the front lens parameter; Output the corrected pixels per degree of viewing angle, pixel density, field of view angle, and physical pixel conversion factor as display calibration parameters.

[0017] Specifically, the calculation process of the calibration module 102 is based on the display optical principle and geometric mathematical relationship. In calculating the pixel density, the horizontal pixel density is obtained by dividing the horizontal resolution by the screen width, and the vertical pixel density is obtained by dividing the vertical resolution by the screen height. The calculation result is usually in units of pixels per millimeter or pixels per inch. The purpose of this step is to determine the pixel distribution characteristics of the display unit 106 in the physical space, providing basic data for subsequent space mapping calculation.

[0018] In calculating the number of pixels per degree of viewing angle, the calibration module 102 first calculates the physical arc length corresponding to a unit viewing angle according to the observation distance, and then converts the physical arc length into the number of pixels in combination with the pixel density. In the specific calculation, the angle is converted into the corresponding physical distance by using the arc length formula, and then the physical distance is converted into the pixel unit by using the pixel density parameter. The accuracy of this parameter directly affects the accuracy of the subsequent defocus processing, because the pixel size of the blur circle needs to be converted based on the number of pixels per degree of viewing angle.

[0019] The calculation of the field of view angle uses the trigonometric relationship. The horizontal field of view angle is calculated by the inverse tangent function based on the ratio of the screen width to the observation distance, and the vertical field of view angle is calculated in the same way based on the screen height. In actual calculation, due to the rectangular characteristics of the display screen, the horizontal and vertical field of view angles need to be calculated respectively to ensure that the defocus processing can accurately cover the complete field of view of the user. The field of view angle parameter determines the boundary definition of the defocus region. A too small field of view angle may result in incomplete defocus effect, and a too large field of view angle may exceed the effective stimulation range.

[0020] The calculation of the physical pixel conversion factor is based on the corresponding relationship between the screen physical size and the resolution, and the conversion coefficients in the horizontal and vertical directions are calculated respectively. This parameter realizes the bidirectional conversion between the physical coordinate system and the pixel coordinate system, so that the physical size of the blur circle calculated based on the geometric optical principle can be accurately mapped to the pixel parameters required by the image processing algorithm. The accuracy of the conversion factor directly affects the accuracy of the defocus simulation. A too large error will cause deviation between the simulation effect and the real optical effect.

[0021] When the system is configured with a front lens, the optical correction process becomes particularly important. The front lens changes the propagation path of light, affecting the actual field of view angle and pixel distribution perceived by the user. The calibration module 102 corrects the calculation result by using optical formulas according to the diopter parameter of the front lens. In the correction process, the magnification or reduction effect of the lens and the possible aberration effect need to be considered. The lens correction ensures that the system can provide accurate defocus stimulation effect under different optical configurations.

[0022] The final output of the display calibration parameters is verified and corrected multiple times to ensure the accuracy and reliability of the parameters. These parameters will be used as input data for the hyperopic defocus parameter modeling module 104 to provide the necessary spatial mapping relationship for accurate defocus image generation, thereby ensuring that the entire amblyopia training system can provide scientific and effective visual stimulation.

[0023] In this embodiment, the hyperopic defocus parameter modeling module 104 is specifically used for: According to the display calibration parameters and pupil data, calculate the image blur parameters in the defocus state; Based on the image blur parameters, defocus processing is performed on the basic training images to generate composite training images with clear center and peripheral hyperopic defocus effect; The composite training images are output to the computing unit.

[0024] Specifically, when calculating the image blur parameters, the hyperopic defocus parameter modeling module 104 first performs accurate calculation of the size of the diffused circle based on the principles of geometric optics. This module receives the display calibration parameters such as the number of pixels per degree and pixel density output by the calibration module 102, as well as the user's pupil diameter data provided by the camera acquisition module 101. During the calculation process, the module uses the similarity triangle geometric relationship to deduce the physical diameter of the diffused circle based on the pre-set hyperopic defocus amount, which is usually set to a value within the range of +0.50D to +2.00D, combined with the pupil diameter parameter. The calculation formula is based on the equivalent focal length model of the human eye optical system, and through the product relationship between the defocus amount and the pupil diameter, the size of the diffused circle formed by the standard point light source on the retina in the defocus state is obtained.

[0025] Subsequently, the module uses the physical pixel conversion factor in the display calibration parameters to convert the calculated diffused circle physical size into a diameter value in pixel units. This conversion process ensures that the geometric optics calculation results can be accurately mapped to the field of digital image processing. Based on the pixelated diffused circle size, the module further determines the core parameters of the Gaussian filter operator, including the size of the Gaussian kernel and the standard deviation value. The side length of the Gaussian kernel is usually set to be slightly larger than the pixel diameter of the diffused circle to ensure complete coverage of the diffused circle energy distribution, and the standard deviation parameter is accurately adjusted according to the diffused circle size to make the energy distribution of the Gaussian function highly consistent with the light intensity distribution in the real defocus state.

[0026] In the defocus processing stage, the module performs regional separation on the basic training image, dividing the image into a central clear region and a peripheral defocus region. The central region is usually defined as the region within a 5-degree range of the central visual angle, which maintains the clarity of the original image and ensures that the user can normally perform the vision training task. The peripheral region is blurred using the calculated Gaussian filter operator to simulate the simulation of hyperopic defocus effect. During the filtering process, the module uses convolution operation to calculate the Gaussian kernel and the image pixels point by point to generate a peripheral region image with progressive blur effect.

[0027] The final generated composite training image has clear functional partition characteristics. The clear image in the central region is used to support the vision training of the weak eye, and the defocus effect in the peripheral region provides the hyperopic stimulus signal required for axial control, achieving the synergistic promotion effect of vision improvement and axial development.

[0028] In the embodiment, the calculation of the image blur parameter in the defocus state according to the display calibration parameter and the pupil data comprises: According to the preset defocus amount and the pupil diameter in the pupil data, the size of the diffused circle formed by the standard point light source on the retina in the defocus state is calculated by geometric optics principle; The display calibration parameter is used to convert the size of the diffused circle into a pixel diameter; According to the pixel diameter, the kernel size and standard deviation of the Gaussian filter operator are determined and output as the image blur parameter.

[0029] Specifically, in the calculation process of the size of the diffused circle, the hyperopic defocus parameter modeling module 104 uses a standard geometric optics model for accurate calculation. This calculation is based on a simplified model of the human eye optical system, which equivalent the complex intraocular optical structure to a single optical system with fixed focal length. In this model, when the parallel light emitted by the standard point light source enters the human eye, in the ideal focusing state, the light rays will converge precisely on a single image point on the retina, forming a clear imaging effect. In the set hyperopic defocus state, due to the insufficient refractive power of the refractive system, the light convergence point is located behind the retina, resulting in a blurred circular light spot, i.e. diffused circle, on the retina plane.

[0030] The calculation of the diameter of the diffused circle is derived using similar triangle geometry. Key parameters involved in the calculation include a preset defocus amount D, usually in diopters, representing the degree of deviation from the ideal focus state to the actual refraction state; the user's pupil diameter d, which is obtained in real time by the camera acquisition module 101 and directly affects the width of the light beam entering the eye; and the defocus distance in the direction of the optical axis, which is converted from the defocus amount by an optical formula. Based on these parameters, the geometric relationship between the light rays at the pupil edge and the optical axis is used to derive the physical diameter calculation formula of the diffused circle through the proportional relationship of similar triangles. This formula fully considers the influence of different pupil sizes on the defocus effect, ensuring the personalized accuracy of the calculation results.

[0031] In the pixel diameter conversion stage, the module establishes a mapping relationship between physical size and pixel units using the physical pixel conversion factor provided by the calibration module 102. During the conversion process, the physical diameter of the diffused circle is first multiplied by the conversion factor to obtain the corresponding pixel diameter value. This conversion takes into account the actual pixel density of the display device and the user's observation distance, ensuring that the physical optical calculation results can be accurately mapped into the pixel coordinate system of digital image processing. The accuracy of the conversion directly affects the accuracy of the subsequent Gaussian filtering process, so this step uses high-precision floating-point operations to minimize the cumulative impact of rounding errors.

[0032] The determination of the Gaussian filter operator parameters is based on the optimized design of the pixel diameter of the diffused circle. The size of the Gaussian kernel is usually set to 1.2 to 1.5 times the pixel diameter of the diffused circle to ensure complete coverage of the energy distribution of the diffused circle while avoiding excessive computational overhead. The calculation of the standard deviation parameter follows the statistical properties of the Gaussian function, so that the full width at half maximum of the Gaussian kernel matches the diameter of the diffused circle, ensuring that the blurred effect after filtering is highly consistent with the actual optical defocus state. In the parameter optimization process, the module also considers the discretization effect of digital sampling and appropriately discretizes the continuous Gaussian function to generate a filter kernel matrix suitable for digital image convolution operations.

[0033] The output image blur parameters include the Gaussian kernel size, standard deviation value, and kernel weight distribution matrix, which together constitute a complete operator description required to accurately simulate the presbyopic defocus effect, providing a scientific and accurate technical foundation for subsequent image processing.

[0034] In this embodiment, the defocus processing of the basis training image based on the image blur parameters to generate a composite training image with a clear center and a presbyopic defocus effect in the periphery includes: separating the basis training image into a central region and a peripheral region; The original clarity of the center region is maintained, and the peripheral region is blurred using a Gaussian filter operator in the image blur parameters to apply a hyperopic defocus effect. The processed center region and peripheral region are recombined to generate the composite training image.

[0035] Specifically, the image region separation process is based on geometric calculation of the angle of view to accurately position. The hyperopic defocus parameter modeling module 104 first converts the angle range of the center region to the corresponding pixel coordinate range according to the number of pixels per degree in the display calibration parameters. The center region is usually defined as a circular region with the image center point as the center and a radius corresponding to a 5-degree angle. The pixel radius of this region is obtained by multiplying the number of pixels per degree by 5 degrees. In actual processing, considering the rectangular pixel array characteristics of digital images, the module uses a distance transformation algorithm to calculate the Euclidean distance of each pixel point to the image center. The pixel points with a distance less than the radius of the center region are marked as the center region, and the remaining pixel points are classified as the peripheral region. This separation method ensures that the center clear region and the peripheral defocus region have a clear boundary definition, avoiding the problem of region overlap or omission in the processing process.

[0036] In the region processing stage, the center region maintains complete original clarity and does not undergo any filtering or blurring processing, ensuring that the key visual information in the amblyopia training task can be clearly presented to the user. This design takes into account the special needs of amblyopia patients for central vision training, ensuring the effectiveness of the training task. For the peripheral region, the module uses two-dimensional Gaussian filter convolution operation for blurring. In the convolution process, the Gaussian filter operator performs sliding window operation at each pixel position in the peripheral region, and the weight distribution of the operator core is weighted and summed with the pixel value at the corresponding position to generate the blurred pixel value. To ensure the accuracy of the boundary processing, the module uses mirror padding method to process the image edges, avoiding the adverse effects of boundary effects on the defocus effect.

[0037] The control of filter strength is based on gradual adjustment of the region position. The module dynamically adjusts the intensity coefficient of the Gaussian filter according to the distance of the pixel point to the boundary of the center region. The farther the pixel point is from the center region, the greater the blurring intensity applied, forming a smooth transition effect from clear to blurred. This gradual design conforms to the physiological characteristics of the human visual system, avoiding abrupt boundary effects and providing a more natural visual stimulation experience. The gradual adjustment algorithm uses a smoothing function for interpolation calculation, ensuring the continuity and smoothness of the blurring intensity change.

[0038] The image synthesis process adopts a pixel-level accurate splicing technology. The module recombines the processed central clear area and the peripheral blurred area according to the original pixel position to generate a complete composite training image. During the synthesis process, special attention is paid to the smooth transition at the junction of the central area and the peripheral area, and the boundary feathering technology is used to reduce the visual jump that may occur. The finally generated composite training image not only maintains the high-definition details of the central area, but also provides accurate hyperopic defocus stimulation in the peripheral area, perfectly realizing the dual functional requirements of vision training and eye axis regulation.

[0039] In the embodiment, the amblyopia training system further comprises: A control unit for receiving the interactive operation instructions of the user and controlling the amblyopia training module to adjust the progress of the training task; A covering unit for covering the user's healthy eye to ensure that the amblyopia eye receives the training stimulation provided by the display unit alone.

[0040] Specifically, the control unit includes various interactive devices, mainly including input devices such as mouse, keyboard and direction handle, for receiving various operation instructions of the user during the training process. The control unit establishes a communication link with the main controller of the amblyopia training system through a standard USB or wireless connection method, ensuring that the operation instructions can be transmitted to the corresponding functional module in real time and accurately. During the training process, the user can perform various interactive operations through the control unit, including starting and pausing the training task, adjusting the difficulty level, setting the training parameters, and querying the training progress, etc. The control unit also supports personalized key configuration, allowing the user to customize shortcut key functions according to personal habits and operational convenience, improving the user experience during the training process.

[0041] After receiving the operation instructions transmitted by the control unit, the amblyopia training module 103 dynamically adjusts the execution parameters of the training task according to the instruction content. The adjustment content includes the selection of the type of training task, the five types of basic tasks of letter recognition, figure matching, stereoscopic target, raster stimulation and contrast sensitivity training built-in the system can be switched according to the user's needs; real-time adjustment of training difficulty, the system supports dynamic adjustment of ten difficulty gradients, and the user can adjust in time according to the training effect and personal ability; control of training duration and frequency, the system can flexibly set the duration of each training and the rest interval according to the user's training plan and physical condition.

[0042] The covering unit adopts a lightweight design, and the weight is controlled to be less than 6 grams, ensuring the comfort during long-time wearing. The covering unit adopts a clamping fixing structure, and can be stably attached to the glasses frame or the head fixing device, avoiding the position deviation due to the slight movement of the head during the training process. The covering material is selected from a professional light shielding material, and has high light shielding rate and good air permeability, which can effectively block the light received by the healthy eye from the outside and the training image information, and can also avoid the stuffiness and discomfort caused by long-time covering. The shape of the covering unit is designed in consideration of the ergonomics principle, so that the visual axis of the healthy eye can be completely blocked when the eyeball of the user rotates in each direction, the interference of the binocular vision information is eliminated, the weak eye can independently receive the composite training image stimulation provided by the display unit, and the pertinence and effectiveness of the training effect are maximized.

[0043] In the embodiment, the amblyopia training system comprises a camera acquisition module, a calibration module, an amblyopia training module, a hyperopic defocus parameter modeling module, a calculation unit and a display unit. The system obtains pupil data of a user through the camera acquisition module, calculates display calibration parameters through the calibration module, and generates a basic training image and a training task through the amblyopia training module. The hyperopic defocus parameter modeling module performs defocus processing on the training image based on the calibration parameters and the pupil data, generates a composite training image with clear center and peripheral hyperopic defocus effect. The calculation unit performs real-time rendering, and the display unit displays the composite training image according to the training task. The present application realizes the cooperation of vision training and eye axis regulation, and provides an effective comprehensive treatment scheme for hyperopic amblyopia patients.

[0044] Please refer to Figure 2 An embodiment of the amblyopia training method in the present application comprises the following steps: 201. Receive the hardware parameters of the display unit through the calibration module, calculate and output the display calibration parameters; In this embodiment, the calibration module receives and processes various hardware parameters of the display unit. For example, when the received parameters are a screen width of 240mm, a screen height of 180mm, a viewing distance of 600mm, and a resolution of 1920×1440 pixels, the calibration module first calculates the pixel density: horizontal pixel density = 1920 pixels ÷ 240mm = 8 pixels / mm, and vertical pixel density = 1440 pixels ÷ 180mm = 8 pixels / mm. Next, it calculates the number of pixels per degree of viewing angle. According to the arc length formula, the physical arc length corresponding to each degree is 600mm × π ÷ 180 ≈ 10.47mm. Therefore, the number of pixels per degree of viewing angle is 8 pixels / mm × 10.47mm ≈ 84 pixels / degree. In the field of view calculation, the horizontal field of view is 2 × arctan(120mm ÷ 600mm) ≈ 22.6 degrees, and the vertical field of view is 2 × arctan(90mm ÷ 600mm) ≈ 17.1 degrees. The physical pixel conversion factor is 8 pixels / mm. When the system is equipped with a +1.5D front lens, the calibration module performs optical correction on the above parameters according to the lens magnification.

[0045] 202. Collect the user's pupil data through the camera acquisition module; In this embodiment, the camera acquisition module 101 monitors the user's eyes in real time. For example, the system identifies the pixel coordinates of the pupil boundary using an edge detection algorithm, and obtains a major axis of 32 pixels and a minor axis of 30 pixels using ellipse fitting, taking an average of 31 pixels. Combining the camera's pixel resolution parameter of 0.1 mm / pixel, the pupil diameter is calculated as 31 pixels × 0.1 mm / pixel = 3.1 mm. To ensure data accuracy, the system continuously acquires 5 frames of images, measuring 3.1 mm, 3.2 mm, 3.3 mm, 3.1 mm, and 3.0 mm respectively, and takes the average value (3.1 + 3.2 + 3.3 + 3.1 + 3.0) ÷ 5 = 3.14 mm as the final pupil data.

[0046] 203. The amblyopia training module generates basic training images and training tasks for the corresponding stages according to the preset training process. In this embodiment, taking a 6-year-old hyperopic amblyopia patient as an example, in the initial stage (week 1), the system selects the letter "E" from the letter recognition task (level 1 difficulty). Based on the 0.5 visual acuity standard, the letter height is calculated as 5 × observation distance (600mm) ÷ 3438 ≈ 0.87mm, but is actually displayed at 20mm to ensure recognizability. This is combined with a +1.00D defocus intensity, and training is conducted for 5 minutes daily. In the advanced stage (week 5), the system selects the graphic matching task (level 5), generating a geometric shape with a diameter of 12mm. The defocus intensity increases weekly using the formula +1.00D + (current week number - 1) × 0.3D = +2.20D, with 10 minutes of training daily. In the consolidation stage (week 10), the system generates a contrast sensitivity training task (level 9), displaying a sinusoidal grating with a contrast ratio of 10^(-9 / 10) ≈ 0.126, maintaining a +3.00D defocus intensity.

[0047] 204. The display calibration parameters and pupil data are received through the farsighted defocus parameter modeling module, the image blur parameters under defocus state are calculated, and the basic training image is defocused based on the image blur parameters to generate a composite training image. In this embodiment, the visual defocus parameter modeling module calculates the circle of confusion based on geometric optics principles. For example, when a pupil diameter of 3.14 mm and a set defocus amount of +2.5D are received, the module calculates the physical diameter of the circle of confusion to be approximately 73.9 mm using similar triangle relationships. This is then converted to a pixel diameter of 591 pixels using a physical pixel conversion factor of 8 pixels / mm. Based on this, the module designs a Gaussian filter operator with a kernel size of 768 pixels and a standard deviation parameter of 126 to ensure the filtering effect matches the actual defocus state. In the image processing stage, the module employs region separation technology, dividing the basic training image into a central sharp region and a peripheral defocus region. Gaussian convolution is applied to the peripheral region to achieve a blurring effect while maintaining the original sharpness of the central 5-degree viewing angle region, ultimately generating a composite image that meets the training requirements.

[0048] 205. The composite training image is rendered in real time by the computing unit, and the image display signal is output. In this embodiment, the computing unit employs a GPU parallel computing architecture to achieve efficient image rendering. For example, for a composite training image with a resolution of 1920×1440, the computing unit utilizes CUDA cores for pixel-level parallel processing, including operations such as color space conversion, gamma correction, and image enhancement. The system maintains a stable 30fps frame rate output to ensure the smoothness of the training process. The rendering pipeline includes stages such as vertex processing, pixel shading, and post-processing, and optimizes graphics algorithms to reduce rendering latency. Finally, a standard image display signal is output through a digital video interface. The signal contains synchronization pulses, color information, and control data to ensure that the display unit can accurately receive and present the rendering results.

[0049] 206、The display unit receives image display signals and training tasks, displays composite training images according to the training tasks, and realizes the synchronous performance of central clear vision training and peripheral hyperopic defocus stimulation; In this embodiment, the display unit simultaneously receives image display signals from the computing unit and training task instructions from the amblyopia training module. For example, when receiving a letter recognition training task, the display unit displays the composite training image in the center of the screen according to the task parameters, with a display duration of 2 seconds and a brightness adjustment to 200 cd / m². During the display process, the letter "E" in the central area remains in high definition to facilitate user identification, and the peripheral area presents a progressive blur effect to provide hyperopic defocus stimulation. The display unit also provides visual feedback according to the training task requirements, such as displaying a green prompt box after correct identification and a red prompt when incorrect, and records the user's reaction time. At the same time, the display unit cooperates with the covering unit to ensure that the stimulation signal only acts on the amblyopic eye, achieving the effect of targeted monocular training.

[0050] 207、The amblyopia training module monitors the training progress in real time, records the training data, and dynamically adjusts the training parameters and process according to the training effect.

[0051] In this embodiment, the amblyopia training module 103 establishes a complete data monitoring and feedback mechanism. For example, the system records the user's training performance data in real time, including task completion, reaction accuracy, average reaction time, and other key indicators. For example, in a 10-minute training, the user completed the letter recognition task 48 times, with a correct rate of 85% and an average reaction time of 1.8 seconds. The module analyzes the training effect according to the preset evaluation standard, and when the correct rate of three consecutive trainings is less than 60%, the system automatically reduces the defocus intensity by 0.25D, or reduces the task difficulty by one level. Conversely, when the correct rate is consistently higher than 90% and the reaction time is shortened, the system will appropriately increase the training difficulty. The module also generates a training report to record the improvement of vision, such as improving from 0.3 to 0.5 after 4 weeks of training, and the system adjusts the subsequent training plan accordingly to ensure continuous optimization of training effect.

[0052] In the embodiment, the display unit hardware parameters are received by the calibration module and the display calibration parameters are calculated, the camera acquisition module acquires the pupil data of the user in real time, and the amblyopia training module generates the basic training images and training tasks of the corresponding stage according to the preset training process. The hyperopic defocus parameter modeling module calculates the image blur parameters based on the display calibration parameters and the pupil data using the geometric optics principle, performs regional defocus processing on the basic training images, and generates composite training images with clear center and peripheral hyperopic defocus effect. The calculation unit uses the GPU parallel architecture to render the composite images in real time, and the display unit presents the rendering results according to the training task parameters, realizing the cooperative performance of central visual training and peripheral defocus stimulation. The amblyopia training module establishes a complete monitoring feedback mechanism, records the training data in real time, dynamically adjusts the defocus intensity and task difficulty according to the user's performance, and ensures the continuous optimization of the training effect. The control method realizes the implementation of the personalized amblyopia training scheme, and provides a scientific and effective visual function improvement way for the patients with hyperopic amblyopia.

[0053] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system or system, unit can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0054] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An amblyopia training system, characterized in that, The amblyopia training system includes: The camera acquisition module is used to collect the user's pupil data; The calibration module is used to calculate display calibration parameters based on the hardware parameters of the display unit; The amblyopia training module is used to generate basic training images and training tasks; The farsighted defocus parameter modeling module is used to receive the display calibration parameters and the pupil data, perform defocus processing on the basic training image, and generate a composite training image. The computing unit is used to perform real-time rendering processing on the composite training image and output an image display signal; The display unit is used to receive the image display signal and the training task, and display the composite training image according to the training task, so as to realize the simultaneous implementation of central clear vision training and peripheral hyperopic defocus stimulation.

2. The amblyopia training system according to claim 1, characterized in that, The hardware parameters include screen width, screen height, viewing distance, resolution, and front lens parameters. The display calibration parameters include the number of pixels per degree of viewing angle, pixel density, field of view, and physical pixel conversion factor.

3. The amblyopia training system according to claim 2, characterized in that, The calibration module is specifically used for: Calculate the pixel density in the horizontal and vertical directions based on the screen width, screen height, and resolution. Calculate the number of pixels per degree of viewing angle based on the observation distance and pixel density; Based on the screen width, screen height, and observation distance, the horizontal and vertical field of view angles are calculated using geometric relationships. Based on the screen width, screen height and resolution, calculate the conversion coefficient between physical size and pixel coordinates as the physical pixel conversion factor; When front lens parameters exist, optical correction is performed on the results of pixel per degree of viewing angle, pixel density, field of view and physical pixel conversion factor based on the front lens parameters. The output corrected pixel density, field of view, and physical pixel conversion factor per degree of viewing angle are used as display calibration parameters.

4. The amblyopia training system according to claim 1, characterized in that, The hyperopia defocus parameter modeling module is specifically used for: Based on the display calibration parameters and pupil data, calculate the image blur parameters under defocus conditions; Based on the image blur parameters, the basic training image is defocused to generate a composite training image with a clear center and a far-sighted defocus effect on the periphery. The composite training image is output to the computing unit.

5. The amblyopia training system according to claim 4, characterized in that, The step of calculating the image blur parameters under defocus conditions based on the display calibration parameters and pupil data includes: Based on the preset defocus amount and the pupil diameter in the pupil data, the size of the circle of confusion formed on the retina by the standard point light source in the defocus state is calculated by using the principle of geometric optics. The size of the blur circle is converted into a pixel diameter using the display calibration parameters. The kernel size and standard deviation of the Gaussian filter operator are determined based on the pixel diameter and output as image blur parameters.

6. The amblyopia training system according to claim 4, characterized in that, The step of defocusing the base training image based on the image blur parameters to generate a composite training image with a sharp center and a peripherally distant defocus effect includes: The basic training image is separated into a central region and a peripheral region; The original sharpness of the central region is kept unchanged, and the Gaussian filter operator in the image blur parameters is used to blur the surrounding region, applying a farsighted defocus effect; The processed central and peripheral regions are then recombined to generate the composite training image.

7. The amblyopia training system according to claim 1, characterized in that, The amblyopia training system also includes: The control unit is used to receive user interaction commands and control the amblyopia training module to adjust the training tasks. A occlusion unit is used to cover the user's healthy eye, ensuring that the amblyopic eye receives the training stimuli provided by the display unit alone.

8. A control method applied to an amblyopia training system as described in any one of claims 1-7, characterized in that, The control method includes: The calibration module receives the hardware parameters of the display unit, calculates and outputs the display calibration parameters. The camera acquisition module collects the user's pupil data. The amblyopia training module generates basic training images and training tasks for the corresponding stages according to a preset training process. The hyperopia defocus parameter modeling module receives the display calibration parameters and pupil data, calculates the image blur parameters under defocus conditions, and performs defocus processing on the basic training image based on the image blur parameters to generate a composite training image. The computing unit performs real-time rendering processing on the composite training image and outputs an image display signal. The display unit receives the image display signal and the training task, and displays the composite training image according to the training task, so as to realize the simultaneous implementation of central clear vision training and peripheral hyperopic defocus stimulation. The amblyopia training module monitors the training progress in real time, records training data, and dynamically adjusts training parameters and processes based on the training effect.

Images