A binocular vision function personalized adjustment training method and system

Abstract

The application discloses a kind of binocular vision function personalized adjustment training method and system, the method is executed by electronic equipment, and the fixation point position of left eye and right eye of user, fixation duration, eye position offset, follow trajectory, blink frequency and fixation reaction time are obtained by camera or infrared eye movement acquisition unit;Variable depth of field visual target with virtual display distance, size, blur degree, contrast, depth level and binocular parallax change is generated by display algorithm, and control display module presents far and near switching, follow and multi-target training task;Then, according to binocular eye movement data and task completion result, the display distance of subsequent training, moving speed, blur change amplitude, display duration and training rhythm are automatically adjusted.The application does not need mechanical lens switching, and can carry out personalized training control according to real-time eye movement performance of user.

Description

Technical Field

[0001] This invention relates to the fields of visual training and eye-tracking interaction technology, specifically to a personalized adjustment training method and system for binocular vision function. Background Technology

[0002] With increased use of electronic display devices, near-field visual activities have significantly increased, leading some users to experience unstable binocular coordination, slow response when switching between near and far vision, fixation point drift, and unstable target tracking. Current binocular vision training typically requires lenses, prisms, physical targets, or specialized training equipment to guide users through convergence, accommodation, tracking, or fixation training by altering external optical conditions. While this method can achieve certain training objectives, the equipment structure is relatively complex, training parameters are not flexible enough, and manual judgment and adjustment based on training performance are usually necessary, making it difficult to adjust training content in a timely manner according to the real-time status of different users.

[0003] Some existing visual training methods based on screens or VR devices can organize training tasks using different patterns, moving targets, or stereoscopic images. However, these methods often play training content according to a fixed program, and the system usually only records whether the task is completed. It does not make sufficient use of eye movement data such as the actual fixation position of the left and right eyes, eye position deviation, tracking trajectory, blink frequency, and reaction time. It is difficult to determine in a timely manner whether the user is stably fixating on the target, whether both eyes are fixating on the same target, whether there is lag during the switching between near and far distances, and whether there is obvious fatigue during the training process.

[0004] Furthermore, conventional training methods for generating distance variations often rely on external lens switching or fixed training procedures, lacking unified control over the distance perception, size, sharpness, depth of field, and movement speed of the training targets. For electronic display platforms such as tablets, computer monitors, VR devices, or dedicated training screens, if display algorithms could be used to directly generate variable depth-of-field targets with varying distance and sharpness, and then combined with eye-tracking results to adjust subsequent training content, the flexibility and personalization of the training process could be improved without adding complex mechanical structures.

[0005] Therefore, it is necessary to propose a personalized training method and system for binocular vision function based on eye tracking and variable depth targets. Summary of the Invention

[0006] The purpose of this invention is to provide a personalized training method and system for binocular vision function, in order to solve the problems of existing training equipment having complex structure, fixed training content, insufficient utilization of real-time eye movement data, and weak personalized adjustment capabilities.

[0007] To achieve the above objectives, the present invention provides a personalized binocular vision function adjustment training method, comprising: Acquire user's binocular eye movement data, which includes at least one of the following: fixation point position of the left and right eyes, fixation duration, eye position deviation, tracking trajectory, blink frequency, and fixation reaction time; A variable depth-of-field target is generated based on training parameters. The variable depth-of-field target is generated by a display algorithm using virtual display distance parameters, target size parameters, blur parameters, contrast parameters, depth-of-field level parameters, and / or binocular parallax parameters. The control display module displays the variable depth target according to a preset training sequence to form a training task of binocular fixation, tracking, and near-far switching; Calculate binocular coordination performance parameters based on the binocular eye movement data and training task completion results; The virtual display distance, near-far switching amplitude, target movement speed, blur change amplitude, display duration, and / or training rhythm for the next round of training are automatically updated based on the binocular coordination performance parameters.

[0008] It should be further explained that the variable depth-of-field target includes a distant target state, a near target state, and a transitional target state between the distant target state and the near target state. The transitional target state is formed by continuously or hierarchically changing the target size, edge blur, contrast, binocular parallax, and / or spatial position.

[0009] The binocular coordination performance parameters include at least one of the following: distance between left and right eye fixation points, difference in fixation duration between left and right eyes, difference in reaction time between near and far eye switching, eye position deviation, tracking trajectory deviation, fixation stabilization time, number of saccades, and task error rate.

[0010] It should be further explained that automatically updating the next round of training based on the binocular coordination performance parameters includes: increasing the difficulty of the next round of training when the distance between the left and right eye fixation points is less than the first threshold and the task completion accuracy reaches the preset requirement; and reducing the difficulty of the next round of training when the distance between the left and right eye fixation points is greater than the second threshold, the reaction time for switching between near and far distances is prolonged, or the task error rate is increased.

[0011] Further explanation is needed: increasing the difficulty of the next round of training includes increasing the range of near-far switching, increasing the target movement speed, shortening the target display duration, and / or increasing the number of multi-target targets; decreasing the difficulty of the next round of training includes decreasing the range of near-far switching, decreasing the target movement speed, extending the target display duration, and / or reducing the number of multi-target targets.

[0012] It should be further explained that the method also includes a fatigue judgment step: generating a training load judgment result based on changes in blink frequency, increased gaze drift, increased task error rate and / or prolonged gaze reaction time; when the training load judgment result meets the preset fatigue conditions, reducing the target movement speed, reducing the blur change amplitude, extending the rest interval or pausing the current training task.

[0013] The present invention also provides a personalized binocular vision function adjustment training system, comprising: The eye-tracking acquisition module is used to acquire the fixation point position, fixation duration, eye position deviation, tracking trajectory, blink frequency and / or fixation reaction time of the user's left and right eyes; The target generation module is used to generate variable depth-of-field targets based on training parameters. The variable depth-of-field targets are generated by a display algorithm using virtual display distance parameters, target size parameters, blur parameters, contrast parameters, depth-of-field level parameters, and / or binocular parallax parameters. The training task control module is used to control the display module to display the variable depth target according to the preset training sequence, so as to form a training task of binocular fixation, tracking and near-far switching. The binocular coordination analysis module is used to calculate binocular coordination performance parameters based on the data acquired by the eye-tracking acquisition module and the training task completion results. The training scheme update module is used to automatically update subsequent training parameters based on the binocular coordination performance parameters.

[0014] It should be further explained that the eye-tracking acquisition module includes a camera acquisition unit and / or an infrared eye-tracking acquisition unit, and is configured to acquire the pupil position, corneal reflection position or line of sight of the left and right eyes respectively, so as to calculate the fixation point position and eye position offset of the left and right eyes.

[0015] It should be further explained that the target generation module is configured to change at least one of the following through a display algorithm: virtual display distance, target size, edge blur, contrast, depth of field level, binocular parallax, and motion path, in order to form a variable depth of field training target that does not require mechanical lens switching.

[0016] The system also includes a user information establishment module and a fatigue judgment module; the user information establishment module is used to record the user's age, refractive status, basic binocular vision function examination results, training goals, and historical training records; the fatigue judgment module is used to generate a training load judgment result based on at least one of blink frequency, gaze drift, task error rate, and gaze reaction time; the training task control module adjusts the training rhythm or pauses training based on the training load judgment result.

[0017] Beneficial Effects: This invention generates variable depth-of-field targets through a display algorithm, eliminating the need for mechanical lenses or external refractive components to switch training conditions. The system can change parameters such as the virtual display distance, size, blur level, contrast, depth of field, and binocular parallax of the targets, allowing users to complete training tasks from far to near, from near to far, from static to dynamic, and from single to multiple targets within a screen or VR environment. This simplifies the device structure and facilitates use on tablets, computer monitors, VR devices, and dedicated training screens.

[0018] This invention collects eye movement data from both eyes during training, not only determining whether the user has completed the task, but also analyzing information such as the distance between the left and right eye fixation points, the difference in fixation duration, eye position deviation, tracking trajectory deviation, the difference in reaction time between near and far distance switching, blink frequency, and error rate. The system can determine whether the user is fixating on the target with both eyes, whether there is lag or instability during near-far distance switching, and whether the training load is too heavy, making the training feedback more objective and detailed.

[0019] This invention can adjust the training content for the next round based on the user's current performance. When the user's binocular coordination is good, the system can increase the range of near-far switching, increase the speed of target movement, shorten the display time, or increase the number of targets. When the user's performance is unstable or the fatigue score increases, the system can reduce the difficulty, extend the display time, reduce the number of targets, or pause the training. Therefore, the training content can change according to the user's eye movement performance and task completion, improving the relevance, continuity, and user experience of the training. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the structure of the personalized binocular vision function adjustment training system of the present invention; Figure 2 This is a flowchart illustrating the personalized binocular vision function adjustment training method of the present invention. Detailed Implementation

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments. It should be understood that the following embodiments are used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention.

[0022] Combination Figure 1 and Figure 2 As shown, this embodiment provides a personalized binocular vision function adjustment training method and system. Figure 1The diagram illustrates the connections between the various functional modules of the system in this embodiment. The system includes a user information establishment module, a historical training record module, a training plan update module, a training task control module, a visual target generation module, a display module, a user binoculars module, an eye movement acquisition module, a binocular coordination analysis module, and a fatigue judgment module. Specifically, the user information establishment module and historical training record module provide basic user information and historical training data to the training plan update module; the training plan update module generates or updates training parameters based on user information, historical training records, binocular coordination analysis results, and fatigue judgment results; the training task control module controls the visual target generation module to generate variable depth-of-field visual targets based on the training parameters and controls the display module to present the training task to the user's eyes; the eye movement acquisition module collects eye movement data such as fixation point position, fixation duration, eye position deviation, tracking trajectory, blink frequency, and fixation reaction time during the training process; the binocular coordination analysis module calculates binocular coordination performance parameters based on the eye movement data and task completion results; and the fatigue judgment module generates a training load judgment result based on the binocular coordination analysis results and training process data, and provides this result to the training plan update module for adjusting subsequent training content.

[0023] Figure 2 The execution flow of the personalized binocular vision function adjustment training method in this embodiment is illustrated. The method first establishes user information and determines initial training parameters, then acquires the user's binocular eye movement data; based on the training parameters, it generates variable depth-of-field targets and presents training tasks such as near-far switching, tracking, and multi-target switching through a display module; during training, it calculates binocular coordination performance parameters based on the binocular eye movement data acquired by the eye movement acquisition module and the task completion results; subsequently, it determines whether the training load meets the fatigue condition. If the fatigue condition is met, the training load is reduced or training is paused before updating the next round of training parameters; if the fatigue condition is not met, the next round of training parameters is updated based on the binocular coordination performance parameters; the updated training parameters are then used for the next round of variable depth-of-field target generation and training task display. Through the above process, the system can personalize the training content according to the user's real-time eye movement performance and task completion status.

[0024] The system described in this embodiment can be deployed on a tablet computer, desktop computer monitor, VR display device, or dedicated training screen. The system includes a user information establishment module, an eye-tracking acquisition module, a visual target generation module, a training task control module, a binocular coordination analysis module, a fatigue judgment module, and a training plan update module. The user information creation module records the user's age, refractive status, basic binocular vision function examination results, training objectives, and historical training records. The eye movement acquisition module can use a regular camera, an infrared camera, or an infrared eye-tracking unit to acquire the pupil position, corneal reflector, gaze direction, fixation point position, fixation duration, eye position deviation, tracking trajectory, blink frequency, and fixation reaction time of the left and right eyes. The target generation module generates variable depth targets with virtual display distance, size, blur level, contrast, depth of field level, binocular parallax, and motion path changes through display algorithms. The training task control module controls the appearance order, display duration, movement speed, near-far switching amplitude, single-target or multi-target presentation mode, and training difficulty level of the targets. The binocular coordination analysis module determines whether both eyes are jointly fixating on a target based on left and right eye movement data and calculates binocular coordination performance parameters. The fatigue judgment module determines the training load based on changes in blink frequency, fixation drift, error rate, and reaction time. The training scheme update module updates subsequent training parameters based on current training results and historical training records.

[0025] Before training begins, the system establishes a user profile and determines initial training parameters based on the user's age, refractive status, basic binocular vision function test results, and training goals. These initial training parameters include the virtual display distance range, near-far switching amplitude, initial target size, initial target contrast, blur variation amplitude, target movement speed, single-round training duration, rest interval, and difficulty level. Among these, the near-far switching amplitude, target movement speed, allowable focal angle error threshold, allowable tracking error threshold, reference reaction time, and fatigue threshold can be determined based on the user's initial test results, equipment calibration accuracy, eye movement sampling frequency, historical training data, or preset age-group training parameters.

[0026] In this embodiment, the training control process is executed by electronic devices. The system generates, displays, and adjusts training targets based on eye-tracking feedback, and outputs training task control parameters and training process data. It does not require making a direct diagnosis of the disease or directly implementing medical treatment as a necessary step.

[0027] After calibrating the user's eye movement acquisition module, it obtains the coordinates of the gaze points of the left and right eyes on the display plane. Let the first... The left eye fixation point at each sampling time was: The right eye fixation point is The target visual center is All coordinates mentioned above are pixel coordinates. To ensure consistency of dimensions across different display devices, the system converts pixel coordinates to visual angle coordinates.

[0028] Let the pixel pitch of the display screen be The unit is meters per pixel, and the distance from the user's eye to the display screen is... The unit is meters, and the center of the display screen is... Then any point The corresponding horizontal and vertical visual angles are:

[0029]

[0030] in, and The unit is radians. This gives the coordinates of the left eye's visual angle. Right eye visual angle coordinates and target visual angle coordinates In small-angle situations, it can also be used. While approximating the original form, the system internally prefers to use the aforementioned arctangent form to improve calculation consistency at different viewing distances.

[0031] To evaluate whether the left and right eyes are both fixating on the same target, the system calculates the distance between the fixation points of the left and right eyes and the binocular target deviation.

[0032] The distance between the left and right eye fixation points is defined as:

[0033] The binocular fusion fixation point is defined as:

[0034] Binocular target deviation is defined as:

[0035] in, and All units are radians. To obtain dimensionless evaluation values, the system sets an allowable angle of view error threshold. The unit is radians, and the normalized co-fixation score is calculated:

[0036] in, It is a dimensionless value between 0 and 1. The larger the value, the higher the degree to which both eyes are focused on the target.

[0037] The co-fixation score for a training task can be obtained by averaging the sampled points over a training task duration.

[0038] in, This represents the number of valid sampling points in this training task.

[0039] For fixation duration, let the allowable fixation range of the target be defined by the target's visual angle coordinates. Centered on, with radius The range of the left eye fixation point. When the left eye fixation point falls within this range, denote the indicator function. ,otherwise When the right eye fixation point is within this range, denote the indicator function. ,otherwise Let the eye-tracking sampling period be... The unit is seconds. The target fixation durations for the left and right eyes are respectively:

[0040]

[0041] The difference in fixation duration between the left and right eyes is:

[0042] in, , and The units are all seconds. This parameter can be used to evaluate whether the left and right eyes maintain a balanced gaze on the same target.

[0043] The target generation module generates variable depth targets based on training parameters. A variable depth target refers to a target that the system does not rely on external mechanical lens switching, but changes the virtual display distance, size, blur level, contrast, depth level and / or binocular parallax of the target through display algorithms, enabling the user to perform near-far switching, tracking and binocular coordination tasks during fixation.

[0044] Let the actual design height of the target be... The unit is meters, and the virtual display distance is... The unit is meters, and the equivalent focal length parameter of the display device is... If the unit is pixels, then the display height of the icon on the screen can be expressed as:

[0045] This formula represents the virtual distance. When the display size decreases, the target size increases; at virtual distance As the display size increases, the target size decreases, thus creating a visual cue of changing distance. If the system uses a fixed display screen, To display the virtual distance parameters in the algorithm; if the system uses a VR device, It can also work with binocular parallax to form spatial depth cues.

[0046] Target sharpness is controlled by a blur parameter. Let the reference sharpness distance for the current task be... The unit is meters, and the virtual display distance is... The unit is meters. The system first calculates the change in virtual distance display:

[0047] in, The unit is The system converts it to the standard deviation of Gaussian blur at the pixel scale:

[0048] in, The unit is pixels. The unit is pixels per meter. (Target image) The image is obtained after Gaussian kernel processing:

[0049] in, The standard deviation is The two-dimensional Gaussian kernel, * denotes convolution operation. In the above formula, the virtual near-far display change is first processed by coefficients. Converted to pixel quantities, therefore there is no need to convert... The issue of dimensions used directly as pixels. When the system uses VR devices or display modules that support stereoscopic display, the target generation module can also generate spatial depth based on binocular parallax.

[0050] Let the user's interpupillary distance be... The unit is meters, and the equivalent focal length of the equipment is... The unit is pixels, and the reference depth is... The virtual display distance is Then the horizontal disparity in the left and right eye images can be defined as:

[0051] in, The unit is pixels. The display center of the target in the left and right eye images are respectively:

[0052] Using the above method, the system can generate training targets at different depth levels without switching mechanical lenses. For ordinary 2D displays, the system can generate variable depth targets simply by changing the size, blur, contrast, and motion path; for VR display devices, binocular parallax parameters can be further superimposed.

[0053] The training task control module displays variable depth-of-field targets according to a preset training sequence. For example, the system can first display a distant target state, then transition to a transitional target state, and finally display a near target state; it can also execute the training task in reverse order, from near to far. Let the initial virtual distance of one near-far switching task be... The virtual distance to the target is The switching duration is Then in time Within this range, the virtual display distance can be generated using a smooth interpolation method:

[0054]

[0055] in, A dimensionless time proportion between 0 and 1. (Using...) The smooth interpolation function allows the target's velocity to change continuously at the start and end of the switching, reducing the interference of abrupt displays on eye-tracking data. The target's position can be set to static, uniform movement, curved movement, or multi-target switching, depending on the task type. For a tracking task, let the target's motion path be... The system calculates the binocular fusion gaze trajectory. With the target trajectory average deviation

[0056] And allow for following error with the target Normalization yields a follower rating:

[0057] in, The unit is radians. It is a dimensionless value between 0 and 1.

[0058] For tasks involving switching between near and far distances, the system records the target stabilization time point after the target switches from the first virtual distance to the second virtual distance. And detect the time points at which the binocular fusion fixation point enters the target's permissible range and remains continuously. The response time for switching between near and far distances is:

[0059] Its unit is seconds. Let the preset reference reaction time be... If the unit is seconds, then the reaction score is:

[0060] in, The value is dimensionless; the shorter the reaction time, the higher the score. Furthermore, the system can separately detect the time points when the left-eye fixation point and the right-eye fixation point enter and remain continuously within the target's permissible range. and The reaction time for near-far switching in the left eye and the reaction time for near-far switching in the right eye were obtained:

[0061] The time difference between the left and right eyes when switching between near and far distances is:

[0062] in, , and The units are all seconds. This parameter can be used to evaluate the synchronicity of the left and right eyes during near-far switching.

[0063] Regarding the task error rate, let the total number of current tasks be... The number of errors is The accuracy rate is:

[0064] in, It is a dimensionless value between 0 and 1.

[0065] The binocular coordination analysis module integrates co-fixation score, following score, response score, and accuracy rate to obtain binocular coordination performance parameters.

[0066] The overall score can be expressed as:

[0067] in, Scoring of binocular coordination performance. , , , These are weighting coefficients, all of which are dimensionless values ​​and satisfy the following:

[0068] When the training task does not include a following task, it can be set as follows: And renormalize the remaining weights; when the training task does not include a reaction task, you can let Then, the remaining weights are renormalized. In this way, a unified scoring framework can be used for different types of training tasks, and each scoring item is a dimensionless value between 0 and 1.

[0069] To further utilize the differences between the left and right eyes, the system can also measure the difference in fixation duration between the two eyes. and the reaction time difference between the left and right eyes when switching between near and far distances As a penalty, the revised overall score is obtained:

[0070] in, For reference fixation duration difference, the unit is seconds; For reference, the response time difference between near and far distances is measured in seconds. and These are dimensionless weighting coefficients; This is the revised dimensionless binocular coordination performance score. The training program update module automatically adjusts the training content for the next round based on the binocular coordination performance parameters.

[0071] Let the current number be... The difficulty of the round training is The range of values ​​is The target performance score is The difficulty of the next round of training can then be updated as follows:

[0072] in, and For dimensionless update coefficients, For fatigue rating, Indicates will Limited to Within this range, the formula states that when the binocular coordination performance score is higher than the target score and the fatigue score is lower, the difficulty of the next round is increased; when the performance score is lower than the target score or the fatigue score is higher, the difficulty of the next round is decreased or maintained. The difficulty level is further mapped to specific training parameters.

[0073] For example, the switching amplitude between near and far distances can be expressed as:

[0074] in, , , All units are .

[0075] The target's movement speed can be expressed as:

[0076] in, , , The units are all radians per second.

[0077] The duration of the target display can be expressed as:

[0078] in, , , The unit for all values ​​is seconds.

[0079] The number of multiple target sights can be expressed as:

[0080] in, To determine the number of multi-target targets in the next round of training. and These are the minimum and maximum values ​​for the number of multi-target sights, respectively; all three are positive integers. This indicates taking the nearest integer. Through the above mapping relationship, the dimensionless parameter of difficulty level can be converted into the range of near / far switching, movement speed, display duration, and number of multi-target targets, each with a definite dimension or value type.

[0081] The fatigue assessment module generates training load assessment results based on changes in blink frequency, gaze drift, task error rate, and gaze reaction time. Let the user's baseline blink frequency be... The current blinking frequency is All units are times / minute; baseline gaze drift is... Current gaze drift is All units are in radians; baseline reaction time is... The current reaction time is All units are seconds; the current task error rate is... , which is a dimensionless value.

[0082] The system can calculate fatigue scores:

[0083] in, , , Let be dimensionless weighting coefficients, and satisfy:

[0084] The above items are all relative changes or error rates, therefore It is a dimensionless numerical value. If greater than the preset fatigue threshold The training task control module reduces the target movement speed, decreases the blur variation, extends the rest interval, or pauses the current training task.

[0085] Specifically, the target movement speed before pausing can be adjusted as follows:

[0086] The fuzzy variation range can be adjusted as follows:

[0087] The rest interval can be adjusted as follows:

[0088] in, , , All are dimensionless adjustment coefficients between 0 and 1. and The units are consistent, both being radians per second; and The units are consistent, all being pixels; and The units are consistent, both being seconds.

[0089] In a specific training process, the system first reads the user profile and determines the initial training level. Then, the eye-tracking acquisition module is invoked for calibration to obtain the mapping relationship between the left and right eye fixation points and the display screen coordinates. The training task control module sends training parameters to the optotype generation module, which then generates the optotypes based on the virtual display distance. Target display height Fuzzy standard deviation Contrast parameters and, if necessary, binocular parallax. A variable depth-of-field target is generated. The display module presents the target according to the training sequence. For example, it first presents a distant, small-sized, high-resolution target, then transitions to a near, large-sized target with different resolutions through a continuous transition, allowing the user to complete a near-far gaze switching task. Next, a target moving along a preset trajectory is presented, allowing the user to complete a binocular tracking task. Finally, multiple targets are presented, requiring the user to complete a target-switching gaze task according to task prompts. Throughout the training process, the eye-tracking acquisition module continuously collects data from the left and right eyes. The binocular coordination analysis module calculates the co-fixation score, tracking score, reaction score, accuracy, and the difference in fixation duration and near-far gaze switching reaction time between the left and right eyes in real time. The training scheme update module updates the training scheme based on the corrected comprehensive score. and fatigue score Update the training level and specific training parameters for the next round.

[0090] When the distance between the left and right eye fixation points is less than the first threshold, the binocular target deviation remains within the allowable range, and the task completion accuracy meets the preset requirements, the system determines that the current training task is performed well and increases the difficulty of the next round of training. This can be manifested by increasing the near-far switching amplitude, increasing the target movement speed, shortening the target display duration, increasing the number of multi-target targets, or increasing the number of depth levels. When the distance between the left and right eye fixation points is greater than the second threshold, the near-far switching reaction time is prolonged, the difference in near-far switching reaction time between the left and right eyes is increased, the difference in fixation duration between the left and right eyes is increased, the tracking trajectory deviation is increased, or the task error rate is increased, the system determines that the current task difficulty is too high and reduces the difficulty of the next round of training. This can be manifested by decreasing the near-far switching amplitude, decreasing the target movement speed, extending the target display duration, decreasing the number of multi-target targets, or decreasing the number of depth levels. When the fatigue score exceeds the preset fatigue threshold, even if the binocular coordination performance score meets the preset requirements, the system will still prioritize reducing the training load or pausing training to avoid continuous fatigue accumulation during the training process.

[0091] In this embodiment, the variable depth-of-field target is not achieved through mechanical lenses or external refractive stimulation switching, but rather generated by a display algorithm that controls the virtual display distance, size, blur level, contrast, depth-of-field levels, binocular parallax, and motion path. The system obtains the user's binocular eye movement performance through an eye-tracking acquisition module, calculates quantifiable binocular coordination performance parameters through a binocular coordination analysis module, and then adjusts the training parameters for the next round through a training scheme update module. These modules can be deployed in the same electronic device or separately in the eye-tracking acquisition device, display device, and control terminal, and interact with each other via wired or wireless communication.

[0092] Through the above implementation methods, this solution can generate training targets with changes in distance, sharpness, size, depth, and motion using the display module without the need for mechanical lens switching. The training parameters can be adjusted based on data such as the user's left and right eye fixation points, fixation stability, eye position deviation, tracking trajectory, blink frequency, and fixation reaction time.

[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A personalized training method for binocular vision function, characterized in that, include: Acquire user's binocular eye movement data, which includes at least one of the following: fixation point position of the left and right eyes, fixation duration, eye position deviation, tracking trajectory, blink frequency, and fixation reaction time; A variable depth-of-field target is generated based on training parameters. The variable depth-of-field target is generated by a display algorithm using virtual display distance parameters, target size parameters, blur parameters, contrast parameters, depth-of-field level parameters, and / or binocular parallax parameters. The control display module displays the variable depth target according to a preset training sequence to form a training task of binocular fixation, tracking, and near-far switching; Calculate binocular coordination performance parameters based on the binocular eye movement data and training task completion results; The virtual display distance, near-far switching amplitude, target movement speed, blur change amplitude, display duration, and / or training rhythm for the next round of training are automatically updated based on the binocular coordination performance parameters.

2. The personalized binocular vision function adjustment training method according to claim 1, characterized in that, The variable depth-of-field target includes a distant target state, a near target state, and a transitional target state between the distant target state and the near target state. The transitional target state is formed by continuously or hierarchically changing the target size, edge blur, contrast, binocular parallax, and / or spatial position.

3. The personalized binocular vision function adjustment training method according to claim 1, characterized in that, The binocular coordination performance parameters include at least one of the following: distance between left and right eye fixation points, difference in fixation duration between left and right eyes, difference in reaction time between near and far eye switching, eye position deviation, tracking trajectory deviation, fixation stabilization time, number of saccades, and task error rate.

4. The personalized binocular vision function adjustment training method according to claim 1, characterized in that, The automatic updating of the next round of training based on the binocular coordination performance parameters includes: increasing the difficulty of the next round of training when the distance between the left and right eye fixation points is less than a first threshold and the task completion accuracy reaches the preset requirement; and reducing the difficulty of the next round of training when the distance between the left and right eye fixation points is greater than a second threshold, the reaction time for switching between near and far distances is prolonged, or the task error rate is increased.

5. The personalized binocular vision function adjustment training method according to claim 4, characterized in that, Increasing the difficulty of the next round of training includes increasing the range of near-far switching, increasing the target movement speed, shortening the target display duration, and / or increasing the number of multi-target targets; decreasing the difficulty of the next round of training includes decreasing the range of near-far switching, decreasing the target movement speed, extending the target display duration, and / or reducing the number of multi-target targets.

6. The personalized binocular vision function adjustment training method according to claim 1, characterized in that, It also includes a fatigue judgment step: generating a training load judgment result based on changes in blink frequency, increased gaze drift, increased task error rate and / or prolonged gaze reaction time; when the training load judgment result meets the preset fatigue conditions, reducing the target movement speed, reducing the blur change amplitude, extending the rest interval or pausing the current training task.

7. A personalized binocular vision function adjustment training system, characterized in that, include: The eye-tracking acquisition module is used to acquire the fixation point position, fixation duration, eye position deviation, tracking trajectory, blink frequency and / or fixation reaction time of the user's left and right eyes; The target generation module is used to generate variable depth-of-field targets based on training parameters. The variable depth-of-field targets are generated by a display algorithm using virtual display distance parameters, target size parameters, blur parameters, contrast parameters, depth-of-field level parameters, and / or binocular parallax parameters. The training task control module is used to control the display module to display the variable depth target according to the preset training sequence, so as to form a training task of binocular fixation, tracking and near-far switching. The binocular coordination analysis module is used to calculate binocular coordination performance parameters based on the data acquired by the eye-tracking acquisition module and the training task completion results. The training scheme update module is used to automatically update subsequent training parameters based on the binocular coordination performance parameters.

8. The personalized binocular vision function adjustment training system according to claim 7, characterized in that, The eye-tracking acquisition module includes a camera acquisition unit and / or an infrared eye-tracking acquisition unit, and is configured to acquire the pupil position, corneal reflection position or gaze direction of the left and right eyes respectively, so as to calculate the fixation point position and eye position offset of the left and right eyes.

9. The personalized binocular vision function adjustment training system according to claim 7, characterized in that, The target generation module is configured to change at least one of the following parameters through a display algorithm: virtual display distance, target size, edge blur, contrast, depth of field level, binocular parallax, and motion path, in order to form a variable depth of field training target that does not require mechanical lens switching.

10. The personalized binocular vision function adjustment training system according to claim 7, characterized in that, It also includes a user information establishment module and a fatigue judgment module; the user information establishment module is used to record the user's age, refractive state, basic binocular vision function examination results, training goals and historical training records; the fatigue judgment module is used to generate a training load judgment result based on at least one of blink frequency, gaze drift, task error rate and gaze reaction time, and the training task control module adjusts the training rhythm or pauses training based on the training load judgment result.

Images