A method and system for near vision functional training

Abstract

The present application relates to the technical field of myopia prevention and control, and discloses a near vision function training method and system, the training method comprising the following steps: constructing a visual training interface containing a multi-dimensional visual function training area on a display terminal; outputting visual training tasks in the training area as 3D visual stimulation pictures in the form of naked-eye stereoscopic through a preset naked-eye 3D stereoscopic rendering algorithm; in the training process, acquiring eye position information of a training object in real time through an eye movement tracking acquisition unit, and dynamically matching and adjusting a three-dimensional image display area; controlling the training content difficulty to gradually increase through a preset training control algorithm, and synchronously receiving touch interaction operations to complete the whole process of visual training; and outputting the training indicators of this time after the training, which are processed by numerical values. The present application overcomes the problem that the existing myopia prevention and control and visual training can only temporarily correct vision and cannot fundamentally improve the core visual function defects related to myopia.

Description

Technical Field

[0001] This invention relates to the technical field of myopia prevention and control, specifically to a method and system for training near-field visual function. Background Technology

[0002] Existing vision training solutions mostly rely on 2D planar training equipment in offline institutions, which is costly in terms of time and money, making it difficult to persist in the long term. Moreover, 2D images cannot achieve accurate binocular stereoscopic stimulation, making it difficult to target and strengthen core visual functions such as binocular fusion and stereoscopic acuity. A few naked-eye 3D training solutions have defects such as the inability of the image to dynamically adjust with the position of the eyeball, the lack of adaptive matching of training difficulty, and the absence of a closed loop of data throughout the process. They cannot be adapted to standardized and personalized deployment in home scenarios, and the training effect is uncontrollable and has poor reproducibility, making it difficult to achieve large-scale promotion and application. Summary of the Invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the technical problem solved by this invention is that existing myopia prevention and vision training methods can only temporarily correct vision and cannot fundamentally improve myopia-related core visual function defects. They also suffer from insufficient training targeting, low stimulation accuracy, poor user compliance, and weak reproducibility of effects. The core issue is how to construct a closed-loop training system to achieve standardized and personalized myopia prevention and vision function correction in the home setting.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a near-field visual function training method, comprising the following steps: constructing a visual training content interface on a display terminal, wherein the visual training content interface is provided with a multi-dimensional visual function training area; outputting the visual training tasks within the multi-dimensional visual function training area in a naked-eye stereoscopic manner using a preset naked-eye 3D stereoscopic rendering algorithm to form a 3D visual stimulation image; during the presentation of the training image, acquiring the eye position information of the training subject in real time through an eye-tracking acquisition unit integrated into the display terminal, and dynamically matching and adjusting the display area of ​​the three-dimensional image according to the eye position information; controlling the training content within the multi-dimensional visual function training area to progressively increase the training difficulty through a preset training control algorithm, synchronously receiving the touch interaction operations of the training subject to complete the entire visual training process; and outputting the numerically processed training indicators after completing all training tasks within the preset training sequence.

[0006] As a preferred embodiment of the near-field visual function training method described in this invention, the multi-dimensional visual function training area includes six targeted training domains: a visual-motor function area, a dynamic visual acuity function area, an eye movement control function area, a visual perception function area, a sensory integration function area, and a binocular fusion function area. Basic visual response training groups are set within the visual-motor function area and the dynamic visual acuity function area. These basic visual response training groups include rapid visual training and hand-eye coordination training. The rapid visual training uses an algorithm to control the first training symbol to randomly appear around the center point of the screen, and the training subject observes the target through peripheral vision and completes a touch-to-eliminate operation. The hand-eye coordination training uses an algorithm to control the rotation direction and speed of the rotating target, gradually increasing the rotation speed with each training round, and receives touch-to-eliminate operations from the training subject on the rotating target.

[0007] As a preferred embodiment of the near-field visual function training method of the present invention, a visual cognitive processing training group is set up within the eye movement control functional area and the visual perception functional area. The visual cognitive processing training group includes visual memory training, sequential memory training, and brain memory training. The visual memory training uses an algorithm to control the random flashing of 1 to 9 graphics or symbols, and the trainee completes the click operation according to the flashing order. The sequential memory training uses an algorithm to present the target in sequence, then the screen goes black, and then the target is randomly shuffled and presented again. The trainee completes the click operation according to the memory order. The brain memory training presents a second training symbol in a quick display manner, requiring the trainee to complete the memorization operation, thereby realizing visual information cognitive processing and memory reconstruction.

[0008] As a preferred embodiment of the near-field visual function training method described in this invention, the stereoscopic vision enhancement training group is set up within the binocular fusion functional area. The stereoscopic vision enhancement training group includes depth training and 3D motion training. The depth training adjusts the inner and outer depth of field of the image using left and right adjustment controls. After the training subject touches a fixed image position, the algorithm gradually increases or decreases the depth parameters to increase the training difficulty. The 3D motion training uses a 3D local parallax algorithm to synthesize moving white dots into a 3D image in real time and display numbers on the white dots. As the training progresses, the number of small balls gradually increases from 1 to 9, and the training subject is required to read the numbers in sequence.

[0009] As a preferred embodiment of the near-field visual function training method described in this invention, the dynamic visual acuity functional area and the binocular fusion functional area are further provided with a visual acuity and fusion enhancement training group, which includes Hart table training, 3D reading training, stereo optotype vision training, and stereo character vision training. The Hart table training generates a 3D alphabet using a 3D synthesis algorithm and randomly changes the letter content, allowing the trainee to quickly count the number of identical letters. The 3D reading training, stereo optotype vision training, and stereo character vision training present content through separate left and right eye viewing or 2D / 3D comparison, gradually reducing the character size and reversing the image depth as training progresses, thereby enhancing binocular fusion function and stereo vision.

[0010] As a preferred embodiment of the near-field visual function training method described in this invention, the naked-eye 3D stereoscopic rendering algorithm includes a 3D interweaving synthesis algorithm and / or a 3D local parallax algorithm, used to generate 3D visual stimulation images with realistic depth of field and left-right eye separation effects; the control logic of the training control algorithm includes automatically transitioning the number of training elements from 1-2 elements at the simple level to a maximum of 9 elements at the complex level, controlling the random appearance of training targets, gradually increasing the target movement speed or rotation speed, independently adjusting the image depth of field, and providing color prompts and sound effects feedback for correct / incorrect operations; the full training content achieves the solidification of long-term visual memory of the training object through repeated targeted stimulation of the optic nerve and visual center.

[0011] As a preferred embodiment of the near vision function training method described in this invention, the training process supports the trainee to wear glasses to perform the full training task, and the training adaptation range is not limited by myopia or astigmatism.

[0012] To address the aforementioned problems, the present invention also provides the following technical solution: a near-field visual function training system, employing the aforementioned near-field visual function training method, comprising a training interface construction module, a naked-eye 3D rendering module, an eye-tracking calibration module, a training control module, an interactive training module, and a result output module integrated within a display terminal. Each module is bidirectionally connected via an internal data bus of the display terminal. The output of the training interface construction module is communicatively connected to the input of the naked-eye 3D rendering module, used to construct a visual training content interface on the display terminal. The multi-dimensional visual function training area is located within the visual training content interface. The output of the naked-eye 3D rendering module is connected to the display unit and interactive training unit of the naked-eye 3D flat panel display terminal. The module's input end is connected for communication, used to output the training tasks within the functional area set to the display unit in a naked-eye stereoscopic manner through a preset naked-eye 3D stereoscopic rendering algorithm; the input end of the eye-tracking calibration module is connected to the eye-tracking acquisition unit, and the output end is connected for communication with the control input end of the naked-eye 3D rendering module, used to dynamically match and adjust the display area of ​​the three-dimensional image based on the real-time acquired eye position information; the training control module is bidirectionally connected to the interactive training module, used to control the training content in each functional area to progressively increase the training difficulty, and collaboratively complete the entire visual training process; the input end of the result output module is connected for communication with both the training control module and the interactive training module, used to record the entire training data and output the quantitative results.

[0013] As a preferred embodiment of the near-field visual function training system described in this invention, the input end of the eye-tracking calibration module is electrically connected to the binocular infrared eye-tracking acquisition unit of the display terminal, and the output end is electrically connected to the parameter control port of the naked-eye 3D rendering module; the eye-tracking calibration module has a built-in eye parameter acquisition circuit and an effective gaze area judgment program, which is used to receive the gaze point coordinates, the angle between the visual axes of the two eyes and the pupil diameter electrical signals output by the eye-tracking acquisition unit, and after processing, outputs the image adjustment electrical signals to the naked-eye 3D rendering module to drive the naked-eye 3D rendering module to synchronously adjust the display coordinates, parallax parameters and scaling ratio of the three-dimensional training image.

[0014] As a preferred embodiment of the near-field visual function training system of the present invention, wherein: the first input terminal of the training control module is electrically connected to the historical data output port of the result output module, the second input terminal is electrically connected to the real-time operation data output port of the interactive training module, the first output terminal is electrically connected to the rendering command input port of the naked-eye 3D rendering module, and the second output terminal is electrically connected to the training data input port of the result output module; the training control module has a built-in difficulty adaptive control program, which is used to generate initial training parameter electrical signals based on historical training data, generate difficulty adjustment electrical signals based on real-time operation data, and synchronously transmit the entire process training data to the result output module for storage and statistics.

[0015] The beneficial effects of this invention are: This invention sets up a multi-dimensional visual function training area within the display terminal and divides it into six targeted training domains. Combined with a training interface construction module, it completes a standardized interface partitioning configuration, which can specifically improve multiple myopia-related visual function defects such as accommodation, convergence, binocular fusion, and eye movement, thereby alleviating ciliary muscle spasm from the root cause.

[0016] The naked-eye 3D rendering module uses a naked-eye 3D stereoscopic rendering algorithm to output stereoscopic training images, which significantly improves the accuracy of binocular stereoscopic visual stimulation compared to traditional 2D planar training. In conjunction with the eye-tracking calibration module, the eye position is collected in real time and the 3D training images are dynamically matched and adjusted to effectively avoid invalid fixation and improve the effective utilization rate of training.

[0017] Meanwhile, the training control module can automatically and adaptively adjust the training difficulty and generate personalized training sequences based on the real-time training data of the training subjects, significantly improving training adaptability and long-term user compliance; the interactive training module forms a closed-loop human-computer interaction, and the result output module realizes the quantitative statistics, storage and export of training data, making the training effect quantifiable, traceable and optimizable.

[0018] The overall solution requires no offline professional equipment and can be implemented in a standardized manner in home settings, solving the technical defects of existing technologies such as poor targeting, uncontrollable effects, low compliance, and difficulty in widespread application. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the near-field visual function training method.

[0021] Figure 2 A schematic diagram illustrating the training and differentiation of multidimensional visual functions.

[0022] Figure 3 This is a schematic diagram of the interface for visual memory and eye movement training. Figure 3 (a) is a schematic diagram of a blank training interface. Figure 3 (b) is a schematic diagram of the training interface marked with red dots.

[0023] Figure 4 This is a schematic diagram of the fast vision-visual-motor training interface. Figure 4 (a) is a schematic diagram of the training interface marked in red. Figure 4 (b) is a schematic diagram of the training interface with letter markings.

[0024] Figure 5 This is a schematic diagram of the interface for depth-of-field training—binocular fusion ability training.

[0025] Figure 6 This is a schematic diagram of the brain memory-visual perception training interface.

[0026] Figure 7 This is a schematic diagram of the hand-eye coordination—dynamic vision training interface. Figure 7 (a) is a schematic diagram of the training interface for a rotating target that combines numbers and graphics. Figure 7 (b) is a schematic diagram of the training interface with different colored graphics.

[0027] Figure 8 This is a schematic diagram of a sequential memory-visual perception training interface. Figure 8 (a) Presents the interface in the target sequence. Figure 8 (b) The interface is presented after the target is shuffled.

[0028] Figure 9 This is a sample of part of the video content used as a reward.

[0029] Figure 10 This is a schematic diagram of the interface for training linear motion—eye movement ability.

[0030] Figure 11 This is a schematic diagram of the 3D motion-eye fusion training interface.

[0031] Figure 12 This is a schematic diagram of the interface for training 3D reading—dynamic vision and fusion ability.

[0032] Figure 13 This is a schematic diagram of the time-visual-sensory integration training interface.

[0033] Figure 14 This is a schematic diagram of the interface for Circular Space Training 1—Visual Motion Training. Figure 14 (a) is the original operation interface of training 1 in the double-circle circular space. Figure 14 (b) is the hand-drawn operation interface of training 1 for a double-circle circular space.

[0034] Figure 15 This is a schematic diagram of the interface for training 2 in a circular space—visual motion training.

[0035] Figure 16 A schematic diagram of the Hart table training 1—dynamic vision training interface.

[0036] Figure 17 A schematic diagram of the Hart table training 2—dynamic vision training interface.

[0037] Figure 18 A schematic diagram of the Hart table training 3—dynamic vision training interface.

[0038] Figure 19 This is a schematic diagram of the interface for Saccades Training 1—Eye Movement Ability Training.

[0039] Figure 20 This is a schematic diagram of the interface for Saccades Training 2—Eye Movement Ability Training. In the diagram, Figure 20 (a) is the first interface form of saccadic training 2. Figure 20 (b) is the second interface form of saccadic training 2.

[0040] Figure 21 This is a schematic diagram of the interface for training spatial movement and eye movement abilities.

[0041] Figure 22 This is a schematic diagram of the half-view letter training interface.

[0042] Figure 23 This is a schematic diagram of the half-field symbol training interface.

[0043] Figure 24 A diagram illustrating the interface specifically designed for left-eye balance training.

[0044] Figure 25 A diagram illustrating the interface specifically designed for right-eye balance training.

[0045] Figure 26 A diagram illustrating the interface specifically designed for children's left eye during balance training.

[0046] Figure 27 A diagram illustrating the dedicated interface for children's right eye balance training.

[0047] Figure 28 A schematic diagram of the dedicated interface for training the left and right eyes using central fixed parallax.

[0048] Figure 29 This is a schematic diagram of a stereoscopic visual acuity training interface.

[0049] Figure 30 This is a schematic diagram of a 3D character vision training interface.

[0050] Figure 31 A diagram illustrating the dedicated interfaces for left and right eye training in binocular vision.

[0051] Figure 32 This is a diagram illustrating the dedicated interfaces for training the left and right eyes in fixational latent strabismus.

[0052] Figure 33 Illustrations of dedicated interfaces for left and right eye training of binocular radial lines.

[0053] Figure 34 This is a schematic diagram of the structure and data flow of each module in the near-field visual function training system. Detailed Implementation

[0054] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0055] Example 1

[0056] Reference Figure 1 , Figure 2 , Figure 4 , Figure 7 , Figure 9 , Figure 14 and Figure 15 As an embodiment of the present invention, a method for training near-field visual function is provided, the specific training process of which is as follows: Figure 1 As shown, it includes the following steps: A visual training content interface is constructed on the display terminal 100, and the visual training content interface is set up with a multi-dimensional visual function training area A.

[0057] Among them, the display terminal 100 is a portable tablet device that integrates a naked-eye 3D display screen, a touch interaction unit, and an eye-tracking acquisition unit 101.

[0058] Using a preset naked-eye 3D stereoscopic rendering algorithm, the visual training tasks in the multi-dimensional visual function training area A are output in a naked-eye stereoscopic manner, forming a 3D visual stimulation image.

[0059] During the presentation of the training image, the eye-tracking acquisition unit 101 integrated into the display terminal 100 acquires the eye position information of the training subject in real time, and dynamically matches and adjusts the display area of ​​the three-dimensional image according to the eye position information.

[0060] The eye position information includes the coordinates of the fixation point, the angle between the visual axes of the two eyes, and the pupil diameter. When the real-time collected eye position information exceeds the preset effective fixation area threshold, the system adjusts the display coordinates, parallax parameters, and scaling ratio of the three-dimensional image in real time to ensure that the core area of ​​the training stimulus image always matches the real-time fixation point of the training subject.

[0061] Furthermore, through a pre-set training control algorithm, the training content within the multi-dimensional visual function training area A is managed to progressively increase in training difficulty, while simultaneously receiving touch interaction operations from the training subjects to complete the entire visual training process. The increase and decrease in training difficulty are automatically adapted based on the real-time training progress of the training subjects.

[0062] After completing all training tasks within the preset training sequence, the system outputs the numerically processed training metrics for this training session. The preset training sequence is a combination of single-session training tasks selected and arranged according to a fixed logic from all training items in this invention based on the visual function baseline and historical training data of the training object. It does not represent all training items recorded in this invention. The numerically processed training metrics refer to the conversion of the training process and completion status into quantifiable and statistically significant numerical data, specifically including training completion rate, accuracy rate of each training item, average reaction time, total training time, visual function improvement dimension score, and difficulty progression. Corresponding visual training reports can be generated simultaneously.

[0063] Specifically, refer to Figure 2 In the diagram, the multi-dimensional visual function training area A includes six targeted training domains: visual-motor function area A1, dynamic visual acuity function area A2, eye movement control function area A3, visual perception function area A4, sensory integration function area A5, and binocular fusion function area A6. These six targeted training domains correspond to the six core visual function dimensions required for myopia prevention and control. The training content in each training domain can be executed independently or in combination according to a preset sequence to form a complete visual function training closed loop.

[0064] The visual-motor functional area A1 and the dynamic visual acuity functional area A2 are equipped with basic visual response training groups, which include rapid visual training, hand-eye coordination training, and circular space training.

[0065] Furthermore, the rapid visual training uses an algorithm to randomly generate the first training symbol F1 around the center point of the screen. The training object observes the target through its surrounding vision and completes the touch cancellation operation. This training corresponds to... Figure 4 ,in Figure 4 (a) Figure 4 (b) Two interface forms for rapid visual training are shown respectively.

[0066] The first training symbol F1 can be any one of dots, letters, numbers, or graphics. This embodiment provides two gradient training modes. Figure 4 (a) A solid red dot is preferred, as the target has no complex semantic interference and can focus on surrounding visual stimuli to the greatest extent. Figure 4 (b) Using uppercase English letters can simultaneously enhance the ability to quickly recognize characters, effectively improve the visual recognition speed of the training subjects, and enhance the coordination of what is seen and touched.

[0067] Furthermore, after correctly completing the touch-to-delete operation, a preset confirmation sound effect and visual prompt are triggered. After a single round of training is completed, the training results for this round are displayed, and touching the center point of the screen will start the next round of training.

[0068] Furthermore, the hand-eye coordination training uses algorithms to control the rotation direction and speed of the rotating target, gradually increasing the rotation speed with each training round, and receiving touch cancellation operations from the training subject on the rotating target.

[0069] This training corresponds to Figure 7 ,in Figure 7 (a) Figure 7 (b) The dynamic rotating interface for hand-eye coordination training is shown. The rotating target includes any one of rotating dots with numbers, letters, or graphics. In this embodiment, a rotating target combining numbers and graphics is preferred. This can simultaneously enhance the ability to recognize dynamic targets and the accuracy of detail visual discrimination while training hand-eye coordination, avoid interference of number sequence cognition on training focus, and fit the training target of dynamic visual acuity.

[0070] The initial rotation speed is preferably set to 30 degrees per second. After every 3 rounds of training with an accuracy rate of no less than 80%, the rotation speed is increased by 15 degrees per second, up to a maximum of 180 degrees per second.

[0071] Furthermore, the circle space training uses an algorithm to generate double or single-circle reference graphics on the screen, and the training object completes continuous drawing operations along the trajectory of the reference graphics through touch operations. This training corresponds to... Figure 14 and Figure 15 ,in Figure 14 The interface for training 1 in the double-circle circular space. Figure 15 This is the operation interface for single-loop circular space training 2. The reference shape includes any one of single-loop circles, double-loop circles, and square closed-loop shapes. In this embodiment, the double-loop closed circle is preferably used as the reference shape for circular space training 1, and the single-loop closed circle is preferably used as the reference shape for circular space training 2. This can accurately constrain the drawing trajectory boundary and effectively enhance the bilateral limb coordination and visual-motor precision control ability of the training object.

[0072] Furthermore, the trainee is required to draw circles within the double circles without exceeding the circle boundaries. Once proficient, the trainee can draw simultaneously with both left and right hands within the left and right double circles. Correct completion of the operation triggers corresponding feedback. The training difficulty can be gradually increased by decreasing the circle diameter and increasing the trajectory movement speed.

[0073] Example 2

[0074] Reference Figure 3 , Figure 6 , Figure 8 , Figure 10 , Figure 19 , Figure 20 and Figure 21 This is the second embodiment of the present invention. In this embodiment, a visual cognitive processing training group is set in the eye movement control functional area A3 and the visual perception functional area A4. The visual cognitive processing training group includes visual memory training, sequential memory training and brain memory training, as well as linear movement training, saccade training and spatial movement training.

[0075] Furthermore, refer to Figure 3 Visual memory training uses an algorithm to control 1 to 9 graphics or symbols to flash randomly. The trainee completes the click operation according to the flashing order. The flashing targets include any one of geometric shapes, numbers, letters, and everyday symbols. In this embodiment, solid color geometric shapes are preferred, which can effectively eliminate the interference of textual semantics and allow the training to focus entirely on strengthening visual memory ability and spatial positioning ability.

[0076] Training starts at the basic level with one graphic or symbol. If the accuracy rate in a single training round is not lower than 90%, the number of graphics or symbols will automatically increase, up to a maximum of 9. The flashing time of graphics or symbols will gradually decrease as the difficulty increases. Correct operation triggers a green visual cue, and incorrect operation triggers a red visual cue.

[0077] Sequential memory training uses an algorithm to present targets in sequence, then the screen goes black. The targets are then randomly shuffled and presented again. The trainee completes the click operations according to the remembered order. This training corresponds to... Figure 8 ,in Figure 8 (a) Presents the interface in the target sequence. Figure 8 (b) The target is presented in a shuffled interface. The target can be any one of Arabic numerals, letter combinations, or graphic combinations. In this embodiment, Arabic numeral arrays are preferred. The memory logic is clear, and the training difficulty can be precisely controlled by the number of numbers. It has strong adaptability.

[0078] Furthermore, training starts at a basic level with 2 objectives, gradually increasing the number of objectives as training progresses, up to a maximum of 9. The objectives are presented sequentially for 100 milliseconds each, with a black screen interval of 1000 to 3000 milliseconds. Correctly completing the sequence triggers a green prompt, while incorrect operation triggers a red prompt and resets the current training round.

[0079] Furthermore, the brain memory training presents a second training symbol through a quick-reading method, requiring trainees to complete a memorization task, thereby achieving visual information cognitive processing and memory reconstruction. This training corresponds to... Figure 6 , Figure 6The interface for displaying the second training symbol is provided. The second training symbol includes any one of words, numbers, or letter combinations. In this embodiment, English words consisting of 3 to 8 letters are preferred. The display has high recognition and the training difficulty can be precisely controlled by the number of characters, which conforms to the industry standard for tachometer training in optometry training.

[0080] Furthermore, the training starts at a basic level of 3 characters or numbers and gradually increases to 8 characters or numbers. The presentation time is shortened with each level of training difficulty. After the dictation is completed, the system automatically checks the results and generates a score for this round of training.

[0081] In this embodiment, linear motion training uses an algorithm to control a moving target to perform horizontal, uniform motion on the screen, and the trainee's eyes work together to track the moving target to complete the training. This training corresponds to... Figure 10 The moving target includes any one of dots, balls, letters, and numbers. In this embodiment, colored letters are preferred, which can accurately guide the eyes to complete smooth tracking movements and effectively improve the ability of the eyes to control movement in the same direction. The movement direction can be set from left to right, from right to left, or alternating between the two directions. The initial movement speed is 50 pixels per second, and the speed increases by 20 pixels per second after every 2 rounds of training, up to a maximum of 300 pixels per second.

[0082] Furthermore, saccade training uses an algorithm to randomly switch the target between the center and edge areas of the screen, and the trainee's eyes rapidly switch between these targets to complete the training. This training references... Figure 19 and Figure 20 ,in Figure 19 For the interface of saccade training 1, Figure 20 middle, Figure 20 (a) Figure 20 (b) shows the two interface forms of saccadic training 2.

[0083] The training targets include any one of letters, numbers, and graphic symbols. This embodiment adopts a phased gradient design. Sagging training 1 preferably uses a combination of uppercase English letters and a central red fixed point, while saccade training 2 uses a combination of numbers and a central red fixed point. This can precisely control the saccade amplitude and simultaneously train the fixation switching speed and the ability to identify different types of targets. During the training process, the target switching interval is gradually shortened from 2000 milliseconds to 500 milliseconds.

[0084] Spatial motion training uses an algorithm to control multiple moving targets to perform random multi-directional movements across the entire screen area. The trainee uses both eyes to jump and track all moving targets to complete the training. This training references... Figure 21The moving targets include any one of the following: circles with numbers, graphics with letters, or moving balls. In this embodiment, a solid black circular target is preferred. The target boundary is clear and there is no extra visual interference. The training starts from the basic level of 1 moving target and gradually increases to 9 targets as the training progresses. The target movement speed is gradually increased from 30 pixels per second to 200 pixels per second.

[0085] Example 3

[0086] Reference Figure 5 , Figure 11 , Figure 28 , Figure 31 , Figure 32 and Figure 33 This is the third embodiment of the present invention. In this embodiment, a stereoscopic vision enhancement training group is set in the binocular fusion functional area A6. The stereoscopic vision enhancement training group includes depth of field training and 3D motion training, as well as central fixed parallax training, fixational latent strabismus training and binocular radiation training.

[0087] In this training method, the depth-of-field training involves adjusting the in-field and out-of-field depths of the image using left and right control levers. After the training subject touches a fixed image location, the algorithm progressively increases or decreases the depth-of-field parameters to enhance the training difficulty. This training corresponds to... Figure 5 It is a training interface with a digital depth scale, which can intuitively display the current depth parameter values.

[0088] The training images can include any one of cartoon images, real-life images, or geometric shapes. In this embodiment, high-contrast cartoon real-life images are preferred, as they provide intuitive depth perception, enhance training interest, and effectively improve the compliance of adolescent trainees.

[0089] Furthermore, during training, external depth-of-field exercises are completed first, followed by internal depth-of-field exercises. During the adjustment process, rapid and repeated clicking of the adjustment controls is prohibited. After a single round of difficulty is achieved, the depth-of-field parameters must be gradually reduced until they reach zero before starting the next round of training. The depth-of-field parameters are adjusted in real time through a 3D interlacing synthesis algorithm, with a parallax adjustment range of ±50 pixels. This enables continuous adjustment, dynamic switching, and depth-of-field reversal of the training image's depth of field.

[0090] Furthermore, 3D motion training uses a 3D local parallax algorithm to synthesize moving white dots into 3D images in real time and display numbers on the white dots. As training progresses, the number of small balls gradually increases from 1 to 9, and the trainees are required to read the numbers in sequence.

[0091] This training corresponds to Figure 11The stereoscopic training targets include any one of the following: a small white ball with numbers, a stereoscopic graphic with letters, or a moving stereoscopic visual target. In this embodiment, a small white stereoscopic ball with Arabic numerals is preferred. It can simulate real spatial motion through 3D stereoscopic presentation, and simultaneously train the stereoscopic fusion ability of both eyes and the ability to recognize number sequences, which is consistent with the visual processing process of real and natural eye use.

[0092] During training, the movement speed of the white dot and the duration of the number display are adjusted step by step with each training round. After the trainee finishes reading, they can confirm the result of this round through touch operation. The system automatically checks the order of numbers and the accuracy of reading.

[0093] Figure 31 It is a binocular vision training interface that can be used to enhance visual discrimination ability under binocular fusion function. During the training process, the stability of binocular collaborative work is improved by displaying line-by-line 2D and 3D mixed display and switching depth of field.

[0094] Furthermore, the central fixed parallax training uses a 3D synthesis algorithm to generate upper and lower split-view triangle images with a central contact point, and the training subject fixates on the central contact point to complete the training. (Refer to...) Figure 28 ,in Figure 28 The top view is the view exclusively for the left eye. Figure 28 The view below is the view exclusive to the right eye. In actual display, the image is presented symmetrically from left to right. In this embodiment, for ease of displaying the content, it is presented vertically. The split view includes any one of the following: vertical split triangles, horizontal split view graphics, or centrally symmetrical graphics. In this embodiment, the vertically symmetrical black solid triangle combination image is preferred. The central fixation point is clear, which can accurately stimulate the binocular central fusion function and stabilize the 3D central fixation ability.

[0095] During training, the trainee is required to fixate on the contact point of the two triangles until the image remains still and the colors are uniform. Once stable, the left and right buttons can be used to switch the displayed content for the left and right eyes and repeat the training.

[0096] Fixational latent strabismus training utilizes a 3D synthesis algorithm to generate a crosshair-shaped split view with a center point and a circle. The training subject maintains a fixed fixation center point to complete the training. This training method references... Figure 32 ,in Figure 32 The top view is the view exclusively for the left eye. Figure 32 The view below is the view exclusive to the right eye.

[0097] The training screen includes any one of the following: a crosshair-shaped split view, a radial split view, or a centrally symmetrical split view. This embodiment preferably uses a crosshair-shaped split view with a fixed central point, which can accurately quantify binocular fixation stability and specifically improve binocular coordination abnormalities caused by fixational latent strabismus. The upper right segment of the crosshair is exclusively viewed by the right eye, and the lower left segment is exclusively viewed by the left eye. During training, the trainee is required to fixate on the central point until the crosshair arms remain stationary and the color is uniform. After stabilization, the left and right buttons can be used to switch the displayed content for the left and right eyes and repeat the training.

[0098] Furthermore, binocular radial line training uses a 3D synthesis algorithm to generate radial split-view images with a center point, and the training subject fixates on the center point to complete the training. This training references... Figure 33 The upper half of the image is viewed exclusively by the left eye, and the lower half is viewed exclusively by the right eye. The training image includes any one of the following: a central radial line graphic, a concentric circle split view graphic, or a grid split view graphic. In this embodiment, a uniform black radial line image with the center point as the origin is preferred, which can stimulate the peripheral fusion function of both eyes in all directions and effectively improve stereoscopic vision sharpness.

[0099] During training, the trainee is required to fixate on the center point until all radiation lines are uniformly black and stationary. After stabilization, the left and right buttons can be used to switch the display content of the left and right eyes and repeat the training.

[0100] Example 4

[0101] Reference Figure 12 , Figure 13 , Figure 16 , Figure 17 , Figure 18 , Figure 22 , Figure 23 , Figure 24 , Figure 25 , Figure 26 , Figure 27 , Figure 29 and Figure 30 In the fourth embodiment of the present invention, a visual acuity and fusion enhancement training group is further provided within the dynamic visual acuity functional area A2 and the binocular fusion functional area A6. This group includes Hart chart training, 3D reading training, stereoscopic visual acuity training, and stereoscopic character visual acuity training, as well as semi-field training and semi-field symbol training. Furthermore, a multi-sensory integration training group is provided within the sensory integration functional area A5. This group includes time vision training, balance training, and children's balance training.

[0102] The Hart table training method generates a 3D alphabet using a 3D synthesis algorithm and randomly changes the letter content, allowing the training object to quickly count the number of identical letters. This training method refers to... Figure 16 , Figure 17 and Figure 18 ,in, Figure 16 This is a schematic diagram of the Hart table training 1—dynamic vision training interface. Figure 17 This is a schematic diagram of the Hart table training 2—dynamic vision training interface. Figure 18 A schematic diagram of the Hart table training 3—dynamic vision training interface.

[0103] The alphabet includes any one of the following: a random uppercase English letter matrix, a number matrix, or an E-sign matrix. This embodiment preferably uses a random uppercase English letter matrix, which has clear statistical logic and allows for precise control of training difficulty through letter size and replacement frequency, effectively improving the reading speed and font detail recognition accuracy of the trainees. Swiping the screen to the right randomly changes the alphabet content, while swiping the screen to the left reduces the letter size to increase training difficulty. The initial letter sign is 1.0, gradually decreasing to 0.1.

[0104] Furthermore, 3D reading training, stereoscopic visual acuity training, and stereoscopic character visual acuity training present content by viewing the left and right eyes separately or by comparing 2D and 3D. As training progresses, the character size is gradually reduced and the depth of field of the image is reversed to enhance binocular fusion function and stereoscopic vision.

[0105] 3D Reading Training Reference Figure 12 The training texts include any one of Chinese story texts, English short articles, and number sequences. In this embodiment, Chinese children's story texts with pinyin are preferred, as they are suitable for the cognitive level of the adolescent trainees and can simulate real daily reading scenarios, allowing the training effect to be directly transferred to daily eye use scenarios.

[0106] Furthermore, a single-eye split-view mode can be set to display text content separately for the amblyopic eye, and the content displayed for the left and right eyes can be switched using a toggle button.

[0107] Stereoscopic visual acuity training reference Figure 29 3D character vision training reference Figure 30 The optotypes include any one of uppercase English letters, E-shaped optotypes, and C-shaped optotypes. In this embodiment, uppercase English letter optotypes are preferred, as they conform to the national standards for visual acuity charts. The optotypes have high recognition and can accurately quantify the improvement in visual discrimination ability.

[0108] Furthermore, it can be set to display 2D and 3D in a line-by-line manner, switch between convex and concave depth of field in a line-by-line manner, and complete at least 4 depth of field inversion operations in each training round. The initial character viewpoint is 1.0, which is gradually reduced to 0.1.

[0109] Reference Figure 22 and Figure 23 Half-field training and half-field symbolic training generate training images with a central target and surrounding masking patterns using algorithms. The training subject fixates on the central target to complete the training. Figure 22 This is a half-field training interface. Figure 23 This is a half-field symbol training interface. The central target can be any one of uppercase English letters, geometric symbols, or numbers. In this embodiment, the central target preferably uses uppercase English letters paired with surrounding diagonal square masking graphics, which can accurately stimulate the oblique neural bundles of the visual system and specifically improve peripheral visual acuity.

[0110] Training starts at a simple level with large targets and gradually reduces the target size as training progresses. Pressing the random key changes the content of the central target.

[0111] Reference Figure 13 The time-visual training uses an algorithm to control the synchronous triggering of rhythmic sound effects and visual cues. The training subject presses the corresponding controls alternately according to the rhythm to complete the training.

[0112] Interactive controls include any one of the following: a press button, a touch slider, or a click icon. In this embodiment, a symmetrical yellow press button is preferred, which is paired with a synchronized rhythm sound effect. This provides strong audiovisual synchronization, can accurately quantify perceived rhythm deviations, and effectively promotes the multi-sensory collaborative integration of vision and hearing.

[0113] The training is divided into two phases. Phase I requires the trainee to complete the click operation within 700 milliseconds after the sound effect is triggered, and Phase II requires the trainee to complete the click operation within 600 milliseconds. A green light indicates a correct operation, and a red light indicates an incorrect operation.

[0114] Balance training for children uses a 3D synthesis algorithm to generate multi-line visual images of letters or symbols, which the trainee reads line by line to complete the training. This training corresponds to... Figure 24 , Figure 25 , Figure 26 and Figure 27 ,in Figure 24 A dedicated interface for balanced training of the left eye. Figure 25 A dedicated interface for right-eye training to balance the learning process. Figure 26 A dedicated interface for children's balance training, designed specifically for the left eye. Figure 27 A dedicated interface for children's right eye balance training.

[0115] The training screen includes any one of the following: a multi-row split-view letter matrix, a split-view symbol matrix, or a split-view number matrix. This embodiment preferably uses a 3x5 matrix of uppercase English letters, which allows for phased training of monocular split-viewing and binocular simultaneous vision, precisely enhancing binocular fusion stability. Training is divided into three stages: monocular split-viewing, binocular simultaneous vision, and multi-column split-viewing. The left and right arrow keys switch the content displayed for the left and right eyes, the random key changes the target content, and swiping left reduces the target size to increase the training difficulty.

[0116] Example 5

[0117] In the fifth embodiment of the present invention, the naked-eye 3D stereoscopic rendering algorithm includes a 3D interweaving synthesis algorithm or a 3D local parallax algorithm, which is used to generate 3D visual stimulation images with realistic depth of field and left and right eye separation effects.

[0118] The 3D interleaving synthesis algorithm adopts pixel-level left and right eye separate view encoding logic, and performs row-by-row pixel interleaving processing on the training content exclusive to the left eye and the right eye. The interleaving rule is that RGB pixels are alternately distributed to the left and right eyes in odd and even rows, so as to achieve crosstalk-free separate view display for the left and right eyes.

[0119] The 3D local parallax algorithm can set differentiated horizontal parallax parameters for a specified rectangular area of ​​the training image. The parallax adjustment range is ±60 pixels, which can realize independent stereoscopic effect rendering of local areas of the image. Both algorithms are mature and commercially available standardized rendering technologies in the field of naked-eye 3D display, and there are no fundamental obstacles to their implementation.

[0120] The management logic of the training control algorithm includes automatically transitioning the number of training elements from 1 to 2 elements at the simple level to a maximum of 9 elements at the complex level, controlling the random appearance of training targets, gradually increasing the target's movement speed or rotation speed, independently adjusting the image depth of field, and providing color prompts and sound effects feedback for correct and incorrect operations.

[0121] The difficulty increase triggering condition of the training control algorithm is as follows: if the accuracy of a single training project is not lower than 80% for 3 consecutive rounds, the difficulty level will be automatically increased by 1; if the accuracy is lower than 60% for 2 consecutive rounds, the difficulty level will be automatically decreased by 1, realizing personalized adaptive matching of training difficulty.

[0122] Furthermore, the full-scale training content achieves long-term visual memory consolidation in the trainees through repetitive targeted stimulation of the optic nerve and visual center. The targeted stimulation process is designed and implemented in strict accordance with internationally recognized laws of visual development and the standardized training theory system for visual function rehabilitation. Through domain-specific and graded visual stimulation, it simultaneously strengthens the information processing capabilities of the visual conduction pathway and the brain's visual center, ensuring that the visual function improvement achieved through training can be maintained long-term after training, without any visual regression under normal eye use scenarios.

[0123] Example 6

[0124] In the sixth embodiment of the present invention, the training process supports the trainee to wear glasses to perform the full training task, and the training adaptation range is not limited by myopia or astigmatism.

[0125] The training process does not require removing glasses and can be completed directly under corrected vision. Preferably, it is suitable for training subjects with myopia and astigmatism with spherical power ranging from -0.50D to -10.00D and cylindrical power not exceeding -6.00D. There is no upper limit on refractive power. The training process is non-invasive, drug-free, and has no side effects. It can be standardized and frequently trained in a home setting.

[0126] Example 7

[0127] Reference Figure 34 This is the seventh embodiment of the present invention. This embodiment also provides a near vision function training system, which adopts the near vision function training method in any of the above embodiments. It includes a training interface construction module 200, a naked-eye 3D rendering module 300, an eye-tracking calibration module 400, a training control module 500, an interactive training module 600, and a result output module 700 integrated in the display terminal 100. Each module is a software function module that is fixed in the central processing unit of the display terminal 100. The two-way communication connection and instruction transmission are realized through the high-speed data bus inside the display terminal 100.

[0128] Among them, the display terminal 100 is a portable tablet device that integrates a naked-eye 3D lenticular lens display screen, a capacitive touch interaction unit, a binocular infrared eye-tracking acquisition unit 101, a data storage unit, a display driving unit and an audio feedback unit, providing the hardware operating platform for this system.

[0129] Furthermore, the output end of the training interface construction module 200 is communicatively connected to the input end of the naked-eye 3D rendering module 300, which is used to construct a standardized visual training content interface on the display terminal 100, with the multi-dimensional visual function training area A set within the visual training content interface; and to complete the partitioning organization and execution entry configuration of the six major targeted training domains.

[0130] The six targeted training domains are: visual-motor functional area A1, dynamic visual acuity functional area A2, eye movement control functional area A3, visual perception functional area A4, sensory integration functional area A5, and binocular fusion functional area A6. The module divides each training domain into an independent display area and touch response area, and configures standardized start, pause, exit, and difficulty switching operation controls, as well as dedicated execution entry points for the corresponding training content.

[0131] The output end of the naked-eye 3D rendering module 300 is communicatively connected to the display unit of the naked-eye 3D flat panel display terminal and the input end of the interactive training module 600. It is used to output the training tasks in the multi-dimensional visual function training area A to the display unit in a naked-eye stereoscopic manner through a preset naked-eye 3D stereoscopic rendering algorithm.

[0132] The preset naked-eye 3D stereoscopic rendering algorithms include a 3D interlacing synthesis algorithm and a 3D local parallax algorithm. The 3D interlacing synthesis algorithm uses a left-eye split-view encoding logic that alternates between odd and even rows of RGB pixels to achieve crosstalk-free split-view display for the left and right eyes. The 3D local parallax algorithm can set horizontal parallax parameters within a range of ±50 pixels for a specified area of ​​the training image to achieve continuous adjustment of the image depth, dynamic switching, and local stereoscopic effect rendering.

[0133] The input of the eye-tracking calibration module 400 is electrically connected to the binocular infrared eye-tracking acquisition unit 101 of the display terminal 100, and the output is electrically connected to the parameter control port of the naked-eye 3D rendering module 300. The eye-tracking calibration module 400 has a built-in eye parameter acquisition circuit and an effective gaze area judgment program. It is used to receive the gaze point coordinates, the angle between the visual axes of the two eyes and the pupil diameter electrical signals output by the eye-tracking acquisition unit 101. After noise removal by the Kalman filter algorithm, it outputs the screen adjustment electrical signal to the naked-eye 3D rendering module 300, driving the naked-eye 3D rendering module 300 to synchronously adjust the display coordinates, parallax parameters and scaling ratio of the three-dimensional training screen.

[0134] The eye position information includes three core parameters: gaze point coordinates, binocular visual axis angle, and pupil diameter. The module presets 80% of the display area in the center of the screen as the effective gaze area threshold. When the gaze point coordinates collected in real time exceed the effective gaze area threshold, the module generates an adjustment command within 100 milliseconds and sends it to the naked-eye 3D rendering module 300. The module simultaneously adjusts the display coordinates, parallax parameters, and image scaling ratio of the three-dimensional image, so that the core area of ​​the training stimulus image always matches the real-time gaze point of the training object.

[0135] Furthermore, the first input terminal of the training control module 500 is electrically connected to the historical data output port of the result output module 700, the second input terminal is electrically connected to the real-time operation data output port of the interactive training module 600, the first output terminal is electrically connected to the rendering command input port of the naked-eye 3D rendering module 300, and the second output terminal is electrically connected to the training data input port of the result output module 700. The training control module 500 has a built-in difficulty adaptive control program, which is used to generate initial training parameter electrical signals based on historical training data, generate difficulty adjustment electrical signals based on real-time operation data, and synchronously transmit the entire process training data to the result output module 700 for storage and statistics.

[0136] The difficulty adaptation rules are as follows: if the accuracy rate of a single training item is not lower than 80% for 3 consecutive rounds, the difficulty level will be automatically increased by 1 level; if the accuracy rate is lower than 60% for 2 consecutive rounds, the difficulty level will be automatically decreased by 1 level. The number of training elements covered by the difficulty adjustment increases from 1 to 9, the target movement or rotation speed increases step by step, the image depth adjustment accuracy increases step by step, and the graphic character size decreases step by step. At the same time, a standardized training process can be generated according to the preset sequence of basic visual response, visual cognitive processing, and stereoscopic enhancement.

[0137] The input terminals of the output module 700 are connected to the training control module 500 and the interactive training module 600 respectively. It is used to record the training data throughout the process and output the quantitative results. It can perform statistical analysis on the training data, generate a report on the visual function change trend of the training object and suggestions for subsequent training optimization. The training data is stored in encrypted form and supports standardized export.

[0138] The entire training process includes training completion rate, accuracy of each training item, average reaction time, total training time, visual function improvement dimension score, and difficulty progression. The training data is stored in the local storage unit of the display terminal 100 using the AES256 symmetric encryption algorithm, and also supports standardized encrypted export via USB and wireless transmission interfaces.

[0139] In summary, combining Figures 1 to 33 During use, after the display terminal 100 is powered on, the training interface construction module 200 first completes the initial construction of the visual training content interface and the partition configuration of the six target training domains, and simultaneously transmits the interface layout and training content data to the naked-eye 3D rendering module 300 to complete the system startup preparation.

[0140] After training starts, the training control module 500 generates a personalized training sequence based on the historical data of the training object and preset rules, and sends the training parameters and rendering instructions to the naked-eye 3D rendering module 300. The rendering module generates naked-eye 3D training images that meet the requirements through built-in algorithms and outputs them to the display unit.

[0141] During training, the eye-tracking calibration module 400 collects the eye position information of the training subject in real time through the eye-tracking acquisition unit 101, and dynamically adjusts the rendering module to adjust the screen parameters to ensure accurate delivery of training stimuli. At the same time, the interactive training module 600 collects the touch operation signals of the training subject in real time and transmits them to the training control module 500 to complete the logical judgment and execute the corresponding visual and audio feedback synchronously, forming a closed-loop training process of training content output, operation acquisition, logical judgment, feedback execution, and difficulty adjustment.

[0142] After the full training task is completed, the results output module 700 summarizes the full-process data from the training control module 500 and the interactive training module 600, completes statistical analysis, report generation, and encrypted storage and export, and realizes the closed-loop operation of the entire myopia prevention and correction training process.

[0143] Example 8

[0144] Reference Figures 1 to 34 This embodiment is used to verify the visual function improvement effect and myopia prevention effect of the near vision function training method described in this invention.

[0145] Specifically, this experiment included 12 adolescent subjects, all of whom completed the full 8-week training program without dropout. They were divided into two groups based on gender: male and female. The basic information of the subjects in each group is as follows: The male group (7 in total) consisted of 2 aged 17, 2 aged 19, 1 aged 20, 1 aged 21, and 1 aged 22, all of whom met the inclusion criteria for the experiment.

[0146] The female group (5 in total) consisted of 1 girl aged 17, 1 girl aged 18, 2 girls aged 19, and 1 girl aged 23, all of whom met the inclusion criteria for the experiment.

[0147] Inclusion criteria: 1. Both eyes have habitual corrected visual acuity ≥9 / 10, normal color vision function, and no organic eye lesions; 2. Able to cooperate in completing the entire process of naked-eye 3D visual training and standardized visual function testing; 3. No other myopia intervention measures such as low-concentration atropine or orthokeratology lenses were used during the experiment.

[0148] Exclusion criteria: Patients with eye diseases such as glaucoma, cataracts, retinal diseases, strabismus, amblyopia, or those with a history of eye surgery or trauma.

[0149] Furthermore, the myopia prevention and correction method described in this invention was used throughout the experiment. Standardized training was performed based on the display terminal 100. The experimental period was 8 weeks, with 4 to 5 training sessions per week and a single training session lasting 18 to 22 minutes.

[0150] The training content strictly corresponds to the six target training domains of the multi-dimensional visual function training area A described in this invention. The core training items are completely consistent with the above embodiments, including: rapid visual 3D training of visual motor function area A1, saccade 3D training of eye movement control function area A3, hand-eye coordination 3D training of dynamic visual acuity function area A2, rhythm 3D training of sensory integration function area A5, brain memory 3D training of visual perception function area A4, and 3D depth training of binocular fusion function area A6.

[0151] Furthermore, during the training process, the eye-tracking acquisition unit 101 integrated into the display terminal 100 acquires the subject's eye position information in real time and dynamically matches and adjusts the display area of ​​the three-dimensional training screen; the training control algorithm automatically adapts to the increase and fallback of training difficulty according to the subject's real-time training completion status, ensuring the targeting and adaptability of the training.

[0152] Specifically, standardized visual function tests were performed on all subjects at the baseline period (before training) and at the end of the experiment (after 8 weeks of training). The testing equipment and operating procedures were standardized throughout the process to ensure data consistency.

[0153] The raw data of the male subjects (7 subjects) are shown in Tables 1 and 2 below.

[0154] Table 1

[0155] As shown in Table 1, none of the four male subjects had organic eye diseases or pathological eye movement abnormalities. Their color vision and developmental levels were age-matched, and their baseline visual function covered low, medium, and high levels, making the experimental sample representative.

[0156] After training, all four subjects showed a positive improvement in their overall visual function scores, with no cases of decline. The average improvement was 18.09 percentage points, with the highest improvement of 25.40 percentage points observed in a 17-year-old male subject with weak baseline visual function. This verifies that the method of this invention has a clear improvement effect on male adolescents with different visual function baselines.

[0157] At baseline, the subjects' accommodation amplitude, convergence near point, stereopsis and other core indicators were all lower than the normal reference values ​​to varying degrees. After training, all core indicators showed significant improvement, proving that the six targeted training domains of the present invention can accurately cover the weak links in the visual function of male adolescents.

[0158] Table 2

[0159] As shown in Table 2, all three male subjects met the characteristics of the suitable population for the training program of this invention and had no contraindications for training. Among them, two subjects had non-dominant hand / foot and one subject had left eye motor dominance, covering the adolescent group with atypical dominant side and further expanding the representativeness of the experimental sample.

[0160] After training, all three subjects showed a significant positive improvement in their overall visual function scores, with an average improvement of 26.73 percentage points and a maximum improvement of 32.70 percentage points. The relative improvement rate reached a maximum of 63.74%, which directly verifies that the method of this invention has a strong effect on improving visual function in male adolescents with weak baseline visual function.

[0161] All subjects successfully completed all the core training programs of this invention without any abandonment or training intolerance, demonstrating that the training program of this invention has excellent adaptability and feasibility for male adolescents with different dominant sides and different ocular dominance.

[0162] The raw data of the female subjects (6 subjects) are shown in Tables 3 and 4 below.

[0163] Table 3

[0164] As shown in Table 3 above, none of the three female subjects had organic eye diseases or pathological eye movement abnormalities. Their color vision and development levels were age-matched, covering a group of adolescent girls aged 18-19. The baseline visual function levels were comprehensive, and the experimental sample was targeted.

[0165] All three subjects showed positive improvement in their overall visual function scores after training, with no cases of decline. The average improvement was 10.77 percentage points. Among them, the 19-year-old female subject with weak baseline visual function showed the highest improvement of 16.90 percentage points, with a relative improvement rate of 26.78%. This verifies that the method of the present invention has a stable effect on improving visual function in female adolescents.

[0166] The fixation tracking ability and saccadic movement ability scores of the three subjects were significantly improved after training, which proves that the eye movement-related training program of the present invention can accurately improve the eye movement control ability of female adolescents and fill the gap in the lack of targeted improvement of visual function of female adolescents in existing training programs.

[0167] Table 4

[0168] As shown in Table 4 above, the two female subjects in this table are 17 and 23 years old, respectively, covering the critical period of myopia progression in adolescents and the period of stable visual function in young adults. This further expands the applicable age range of the method of this invention and makes the sample more representative.

[0169] Both subjects showed positive improvement in their overall visual function scores after training. Even the 17-year-old female subject, who had a high baseline visual function level, achieved an improvement of 9.60 percentage points. This demonstrates that the method of the present invention can not only solve the problem of weak visual function, but also further enhance normal visual function, making it applicable to a wider range of scenarios.

[0170] Both subjects successfully completed all core training programs. No eye discomfort or adverse events occurred during the experiment, proving that the training program of the present invention has excellent safety and acceptability for female adolescents of different ages and can be consistently implemented in the long term.

[0171] The results for all 12 subjects are shown in Table 5 below.

[0172] Table 5

[0173] As shown in Table 5, all 12 subjects in the sample achieved a 100% positive improvement in their comprehensive visual function scores after training, with an average improvement of 16.65 percentage points and an average relative improvement rate of 28.91%. This statistically verifies that the visual function improvement effect of the method of this invention is universal and stable, and there are no cases of effect failure due to individual differences.

[0174] The method of this invention has a clear improvement effect on both male and female adolescents. The average relative improvement rate of the male group is 39.07% and that of the female group is 15.73%, which proves that the solution has a more prominent improvement effect on the group with low baseline visual function level and can cover the myopia prevention and control needs of adolescents of different genders and different visual function levels.

[0175] The highest improvement in the entire sample reached 32.70 percentage points. Even subjects with high baseline visual function achieved an improvement of more than 5.40 percentage points, proving that the method of the present invention can not only solve the problem of weak visual function, but also further enhance normal visual function, which is superior to the existing training programs that only target abnormal visual function.

[0176] Both sets of data showed no cases of negative improvement and no records of adverse training events, proving that the method of the present invention has extremely high safety and effectiveness, and can be promoted and applied as a standardized and universal solution for the prevention and control of myopia in adolescents.

[0177] In summary, based on the complete raw data from an 8-week standardized experiment involving 12 subjects, the near-field visual function training method described in this invention can be clearly verified to possess the following distinct technical effects: First, it can comprehensively and significantly improve the core visual function of adolescents, solving the core deficiency of existing myopia prevention and control programs that only focus on refractive correction and lack multidimensional visual function system training. It has a stable and effective training effect on adolescents of different genders, ages and visual function foundations.

[0178] Secondly, it can effectively slow down the progression of myopia, achieving root-cause prevention and control of myopia from the perspective of improving the underlying capabilities of the visual system. By targeting and strengthening core visual functions such as accommodation, convergence, stereopsis, and eye movement, it can relieve ciliary muscle spasm, control the trend of axial elongation, and avoid the problem of rapid vision rebound after discontinuation of traditional optical / drug intervention programs.

[0179] Third, it has strong training targeting, excellent adaptability and safety. It achieves personalized difficulty adaptation through training control algorithms, and the average training compliance of subjects reached 92.8%. No adverse events occurred during the experiment. It can be deployed in a standardized and high-frequency manner in home scenarios, solving the pain points of traditional offline visual training scenarios being limited, costly and difficult to persist in the long term.

[0180] Fourth, the training effect is quantifiable and reproducible. All technical effects are supported by complete original experimental data for gender and individual, which directly correspond to the core technical features of this invention, such as the six targeted training domains, naked-eye 3D stereoscopic rendering, and eye-tracking dynamic matching. This proves that the technical solution of this invention can stably achieve the expected effects of visual function improvement and myopia prevention and control, and has clear practicality and inventiveness.

[0181] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A near-field visual function training system, employing a near-field visual function training method, characterized in that: Includes the following steps: A visual training content interface is constructed on a display terminal (100), wherein the visual training content interface is provided with a multi-dimensional visual function training area (A). Using a preset naked-eye 3D stereoscopic rendering algorithm, the visual training tasks in the multi-dimensional visual function training area (A) are output in naked-eye stereoscopic mode to form a 3D visual stimulation image. During the training screen presentation process, the eye position information of the training object is acquired in real time through the eye tracking acquisition unit (101) integrated in the display terminal (100), and the display area of ​​the three-dimensional image is dynamically matched and adjusted according to the eye position information. Eye position information includes fixation point coordinates, the angle between the visual axes of both eyes, and pupil diameter; Through a preset training control algorithm, the training content in the multi-dimensional visual function training area (A) is controlled to gradually increase the training difficulty, and the touch interaction operation of the training object is received simultaneously to complete the entire process of visual training. After completing all training tasks within the preset training sequence, output the numerically processed training metrics for this training session. The multidimensional visual function training area (A) includes six major targeted training domains: visual-motor function area (A1), dynamic visual acuity function area (A2), eye movement control function area (A3), visual perception function area (A4), sensory integration function area (A5), and binocular fusion function area (A6). Basic visual response training groups are set up in the visual-motor functional area (A1) and the dynamic visual acuity functional area (A2), which include rapid visual training and hand-eye coordination training. The visual cognitive processing training group is set up in the eye movement control functional area (A3) and the visual perception functional area (A4), which includes visual memory training, sequential memory training and brain memory training. The binocular fusion functional area (A6) is equipped with a stereoscopic vision enhancement training group, which includes depth of field training and 3D motion training. The dynamic visual acuity functional area (A2) and binocular fusion functional area (A6) are further equipped with visual acuity and fusion enhancement training groups, which include Hart table training, 3D reading training, stereo chart vision training and stereo character vision training.

2. The near-field visual function training system as described in claim 1, characterized in that: The rapid visual training uses an algorithm to control the first training symbol (F1) to appear randomly around the center point of the screen. The training object observes the target through the surrounding vision and completes the touch cancellation operation. The hand-eye coordination training uses an algorithm to control the rotation direction and speed of the rotating target, gradually increasing the rotation speed with each training round, and receiving touch cancellation operations from the training subject on the rotating target.

3. The near-field visual function training system as described in claim 2, characterized in that: The visual memory training uses an algorithm to control the random flashing of 1 to 9 graphics or symbols, and the training subject completes the click operation according to the flashing order. The sequential memory training uses an algorithm to present the target in sequence, then the screen goes black, and then the target is randomly shuffled and presented again. The training subject completes the click operation according to the memory order. The brain memory training presents the second training symbol through a quick display method, requiring the trainee to complete the dictation operation, thereby realizing visual information cognitive processing and memory reconstruction.

4. The near-field visual function training system as described in claim 2 or 3, characterized in that: The depth of field training is achieved by adjusting the inner and outer depth of field of the image using left and right adjustment controls. After the training object touches a fixed image position, the algorithm gradually increases or decreases the depth of field parameters to increase the training difficulty. The 3D motion training uses a 3D local parallax algorithm to synthesize moving white dots into 3D images in real time and display numbers on the white dots. As the training progresses, the number of small balls gradually increases from 1 to 9, and the trainees are required to read the numbers in sequence.

5. The near-field visual function training system as described in claim 4, characterized in that: The Hart table training uses a 3D synthesis algorithm to generate a 3D alphabet and randomly change the letter content, allowing the training object to quickly count the number of identical letters. The 3D reading training, stereoscopic visual acuity training, and stereoscopic character visual acuity training present content through separate left and right eye vision or 2D / 3D comparison. As the training progresses, the character size is gradually reduced and the depth of field of the image is reversed to enhance binocular fusion function and stereoscopic vision.

6. The near-field visual function training system as described in claim 5, characterized in that: The naked-eye 3D stereoscopic rendering algorithm includes a 3D interweaving synthesis algorithm and / or a 3D local parallax algorithm, which are used to generate 3D visual stimulation images with realistic depth of field and left and right eye separation effects. The control logic of the training control algorithm includes automatically transitioning the number of training elements from 1 to 2 elements at the simple level to a maximum of 9 elements at the complex level, controlling the random appearance of training targets, gradually increasing the target's movement speed or rotation speed, independently adjusting the image depth of field, and providing color prompts and sound effects feedback for correct / incorrect operations. The full training content achieves long-term visual memory consolidation in the trainees through repeated targeted stimulation of the optic nerve and visual center.

7. The near-field visual function training system as described in any one of claims 1 to 3, 5 and 6, characterized in that: During training, trainees can wear glasses to perform the full training tasks, and there are no restrictions on astigmatism or myopia.

8. The near-field visual function training system as described in claim 1, characterized in that: The system includes a training interface construction module (200), a naked-eye 3D rendering module (300), an eye-tracking calibration module (400), a training control module (500), an interactive training module (600), and a result output module (700) integrated within the display terminal (100). Each module is connected to the display terminal (100) via a bidirectional communication bus. The output end of the training interface construction module (200) is communicatively connected to the input end of the naked-eye 3D rendering module (300) for constructing a visual training content interface on the display terminal (100). The multi-dimensional visual function training area (A) is set within the visual training content interface. The output end of the naked-eye 3D rendering module (300) is communicatively connected to the display unit of the naked-eye 3D flat panel display terminal and the input end of the interactive training module (600), and is used to output the training tasks in the functional area set to the display unit in naked-eye stereoscopic mode through a preset naked-eye 3D stereoscopic rendering algorithm. The input end of the eye-tracking calibration module (400) is connected to the eye-tracking acquisition unit, and the output end is connected to the control input end of the naked-eye 3D rendering module (300) for dynamic matching and adjustment of the display area of ​​the three-dimensional image based on the real-time eye position information. The training control module (500) and the interactive training module (600) are bidirectionally connected and used to control the training content in each functional area to progressively increase the training difficulty and collaboratively complete the entire process of visual training. The input terminal of the result output module (700) is connected to the training control module (500) and the interactive training module (600) respectively, and is used to record the training data of the whole process and output the quantization results.

9. The near-field visual function training system as described in claim 8, characterized in that: The input of the eye-tracking calibration module (400) is electrically connected to the binocular infrared eye-tracking acquisition unit (101) of the display terminal (100), and the output is electrically connected to the parameter control port of the naked-eye 3D rendering module (300). The eye-tracking calibration module (400) has a built-in eye parameter acquisition circuit and an effective gaze area judgment program. It is used to receive the gaze point coordinates, the angle between the visual axes of the two eyes and the pupil diameter electrical signals output by the eye-tracking acquisition unit (101). After processing, it outputs the screen adjustment electrical signals to the naked-eye 3D rendering module (300) to drive the naked-eye 3D rendering module (300) to synchronously adjust the display coordinates, parallax parameters and scaling ratio of the three-dimensional training screen.

10. The near-field visual function training system as described in claim 9, characterized in that: The first input terminal of the training control module (500) is electrically connected to the historical data output port of the result output module (700), the second input terminal is electrically connected to the real-time operation data output port of the interactive training module (600), the first output terminal is electrically connected to the rendering instruction input port of the naked-eye 3D rendering module (300), and the second output terminal is electrically connected to the training data input port of the result output module (700). The training control module (500) has a built-in difficulty adaptive control program, which is used to generate initial training parameter electrical signals based on historical training data, generate difficulty adjustment electrical signals based on real-time operation data, and synchronously transmit the entire process training data to the result output module (700) for storage and statistics.

Images