VR-based binocular coordination visual training system and control method
By using a VR-based binocular coordination vision training system, which employs separate eye rendering and dynamic blur control, the problem of binocular competitive inhibition disorder after monocular vision correction surgery has been solved. This system has improved visual adaptability in near-far scene switching for presbyopic individuals and significantly improved visual acuity and inhibition relationship.
Patent Information
- Application Number
- CN202511084386.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-04
- Publication Date
- 2025-11-14
AI Technical Summary
Existing presbyopia training techniques cannot effectively address the binocular competitive inhibition disorder after monocular vision correction surgery, especially when switching between near and far scenes, resulting in insufficient visual adaptation. Furthermore, current VR technology lacks dynamic training programs specifically for presbyopic individuals.
A VR-based binocular coordination vision training system is adopted. The left and right eye fields of vision are physically isolated through the VR binocular display module, and the far and near scene content is rendered simultaneously using the separate eye independent rendering engine. Combined with the dynamic blur controller and voice prompt module, the system realizes the gradual blurring and sharpening switching between the dominant and non-dominant eyes. With the help of the eye tracking module and the central control module, the training rhythm is dynamically adjusted.
It significantly improves binocular visual coordination, corrects binocular inhibition abnormalities, enhances near-far visual adaptation, improves training accuracy and comfort, and strengthens the user's visual adaptation ability.
Smart Images
Figure CN120938780A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to virtual reality (VR) technology, and more specifically to a VR-based binocular coordination visual training system and control method. Background Technology
[0002] Presbyopia is an age-related physiological visual impairment primarily caused by decreased lens elasticity, reduced ciliary muscle strength, and structural changes in the ciliary band and sclera, leading to a significant reduction in the eye's dynamic accommodative range. When an individual's accommodative ability declines to a critical level, even with optimal distance correction, their near vision may still be insufficient for daily near-vision needs, becoming a common visual health problem faced by the elderly.
[0003] Presbyopia not only manifests as a decline in accommodative ability, but may also lead to a decrease in binocular integration ability due to an imbalance between accommodation and fusion, which in turn can cause symptoms such as eye strain, reading difficulties, and weakened stereoscopic perception.
[0004] Currently, most refractive surgical interventions for presbyopia (such as corneal laser surgery or lens replacement) employ a monovision design: the dominant eye is fully corrected to ensure clear distance vision, while the non-dominant eye retains an appropriate amount of myopia to maintain near vision. Although this approach can partially address the dual needs of both distance and near vision, a certain percentage of patients experience binocular vision coordination disorders post-surgery, specifically manifested as:
[0005] 1. Distance vision interference: Blurry images of the non-dominant eye cannot be effectively suppressed by the dominant eye, and the brain receives both clear and blurry signals at the same time, resulting in visual double vision;
[0006] 2. Near vision interference: The far-focus image of the dominant eye competes with the clear image of the near-focus eye, reducing the quality of near vision;
[0007] 3. Abnormal inhibition relationship: In extreme cases, inhibition reversal occurs (the non-dominant eye inhibits the dominant eye when viewing distance, and the dominant eye inhibits the non-dominant eye when viewing near objects), resulting in a simultaneous decrease in both distance and near vision. This type of visual interference severely reduces the user's quality of life and surgical satisfaction.
[0008] Existing visual function training techniques have significant limitations:
[0009] · Mainstream methods (such as chromatic aberration 3D and polarized light 3D technology) are mainly designed for children and adolescents with amblyopia and binocular vision function reconstruction, and cannot adapt to the physiological characteristics of presbyopia.
[0010] · The training scenarios are static and lack dynamic simulation of near-far vision switching, which is out of touch with the daily vision needs of presbyopic users.
[0011] · Precise binocular inhibition control cannot be achieved, and it is particularly difficult to simulate the dominant / non-dominant eye competition relationship unique to monocular vision surgery.
[0012] · The interactivity is insufficient, and the training content does not closely match the real visual scene.
[0013] Although virtual reality (VR) technology has been applied in the field of amblyopia rehabilitation, existing solutions have not yet proposed a systematic solution for the specific visual function defects (such as bidirectional inhibition imbalance and dynamic accommodation conflict) after presbyopia monocular vision design surgery.
[0014] Therefore, there is an urgent need to develop a visual function training system specifically for presbyopic individuals to address the core issue of binocular competitive inhibition disorder after monocular vision surgery and improve visual adaptability during near-far scene switching. Summary of the Invention
[0015] The technical problem to be solved by this invention is to propose a VR-based binocular coordination visual training system and control method to solve the core problem of binocular competitive inhibition disorder after monocular vision correction and improve visual adaptability when switching between far and near scenes.
[0016] The technical solution adopted by this invention to solve its technical problem is:
[0017] A VR-based binocular coordination vision training system includes:
[0018] The VR binocular display module physically isolates the display areas of the left and right eye fields of view through its optical lens barrels;
[0019] The dual-eye independent rendering engine supports simultaneous rendering of scene images containing both distant and near-field content in the display area of both eyes' field of vision, and independently renders the blurring effect of distant and near-field content in the scene images of both eyes.
[0020] The dynamic blur controller, invoked by the eye-independent rendering engine, dynamically controls the progressive blurring process and effect of the blurred object. Specifically, it independently adjusts the Gaussian blur shader radius of the image displayed in the visual field of a single eye according to a set value. The progressive blurring of the dynamic blur controller is achieved through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V is as follows:
[0021] When V∈[0.7,1.0], R=4×(1.0-V) / 0.3;
[0022] When V∈[0.3,0.7), R=4+6×(0.7-V) / 0.4;
[0023] R: Gaussian blur radius (Blur Radius), in pixels (px);
[0024] V: Visual Acuity, normalized value range 0.3 to 1.0;
[0025] Wherein, target sharpness V is a standardized display sharpness parameter that characterizes the recognizability of target images in a VR scene. V = 1.0 represents a blur-free state, and V = 0.7 represents the minimum recognizable threshold, as detailed below:
[0026] V=1.0: The target is completely clear (no blurring).
[0027] V = 0.7: The target has reached the minimum identifiable threshold (ISO 8596:2017 standard);
[0028] V = 0.3: The target is severely blurred (maximum blur radius 10px);
[0029] The blur gradient of the dynamic blur controller is 0.1 to 0.25D / step, and blurring stops when the standardized target sharpness V ≤ 0.7;
[0030] The user information entry module is used to enter the user's age and set the user's dominant eye.
[0031] The voice prompt module is used to issue voice prompts to the user before the gradual blurring begins, guiding the user to adjust their attention to the blurred object;
[0032] An eye-tracking module is used to monitor eye-tracking latency in real time;
[0033] The central control module is used to read user information, set the initial fuzzy object based on the user information, control the switching of fuzzy objects, control the duration of progressive fuzzing, and control the content and timing of language prompts.
[0034] Further optimization of the technical solution: the resolution of the scene image is ≥4K, the near-field content refers to the image with a viewing distance of 40cm and includes dynamic texture details; the far-field content refers to the image with a viewing distance of 5m and above.
[0035] In a further optimized technical solution, the central control module retrieves the user's dominant eye information from the user information input module and sets the initial fuzzy object of the dynamic fuzzy controller based on the user's dominant eye information.
[0036] To further optimize the technical solution, the central control module retrieves the user's age information from the user information input module to determine the duration of the progressive fuzzification of the dynamic fuzzy controller. The determination rule is as follows: when the user's age is ≤55 years old, it is set to 3 seconds; when the user's age is >55 years old, it is set to 5 seconds.
[0037] In a further optimized technical solution, during the progressive blurring process of the dynamic blur controller, the eye-tracking module monitors the eye-tracking delay in real time. If the delay is >500ms, it sends a signal to the central control module, which then controls the dynamic blur controller to extend the blurring duration by 0.5 seconds.
[0038] Further optimizing the technical solution, the central control module controls the independent rendering engine to alternately render the progressive blurring effect of far and near content in the display area of the VR dual-eye display module, and the number of cycles is controlled to be 10 to 15.
[0039] Further optimization of the technical solution: the user's dominant eye is defined as the farsighted visual dominant eye.
[0040] A control method for a VR-based binocular coordination vision training system, characterized in that the control method includes the following steps:
[0041] S1. User Information Entry: The user information entry module collects and enters the user's age and sets the user's dominant eye.
[0042] S2. Device initialization settings: The central control module controls the eye-separated independent rendering engine to generate scene images containing far and near content in the display areas of both eyes, and controls the dynamic blur controller to initialize settings so that the far and near content of the scene images in the display areas of both eyes is clear.
[0043] S3. Long-range image competition suppression: This includes the following steps:
[0044] S31. The central control module retrieves the user's primary eye information from the user information input module, and based on the user's primary eye information, sets the distant content in the non-primary eye field of view as the initial blurred object, and transmits the initial blurred object information to the split-eye independent rendering engine.
[0045] S32. The central control module first sends instruction information to the voice prompt module, the voice prompt module responds and sends a first voice instruction to the user, prompting the user to continue to pay attention to the distant content in the scene image;
[0046] S33. The independent rendering engine for each eye calls the dynamic blur controller to progressively blur distant content within the non-primary eye's field of vision, while maintaining clarity of distant content within the primary eye's field of vision. The blur gradient is 0.1–0.25D / step, until the standardized target sharpness V ≤ 0.7. The duration of the progressive blurring is set based on user information, with the following rules: 3 seconds for users aged ≤ 55, and 5 seconds for users aged > 55. During the progressive blurring process, eye-tracking delay is monitored in real time; if the delay > 500ms, the blurring duration is extended by 0.5 seconds.
[0047] The progressive blurring of the dynamic blur controller is achieved through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V is as follows:
[0048] When V∈[0.7,1.0], R=4×(1.0-V) / 0.3;
[0049] When V∈[0.3,0.7), R=4+6×(0.7-V) / 0.4;
[0050] R: Gaussian blur radius (Blur Radius), in pixels (px);
[0051] V: Visual Acuity, normalized value range 0.3 to 1.0;
[0052] Wherein, target sharpness V is a standardized display sharpness parameter that characterizes the recognizability of target images in a VR scene. V = 1.0 represents a blur-free state, and V = 0.7 represents the minimum recognizable threshold, as detailed below:
[0053] V=1.0: The target is completely clear (no blurring).
[0054] V = 0.7: The target has reached the minimum identifiable threshold (ISO 8596:2017 standard);
[0055] V = 0.3: The target is severely blurred (maximum blur radius 10px);
[0056] S34. Restore the distant view content in the non-dominant eye's field of vision to full clarity;
[0057] S4. Close-range image competition suppression: The specific steps are as follows:
[0058] S41. The central control module sends a blur object switching command to the split-eye independent rendering engine. The split-eye independent rendering engine responds to the command and switches the blur object to the near-field content within the main eye's field of view.
[0059] S42. The central control module first sends an instruction message to the voice prompt module, the voice prompt module responds and sends a second voice instruction to the user, prompting the user to continue to pay attention to the close-up content in the scene image;
[0060] The independent rendering engine for each eye calls the dynamic blur controller to progressively blur the near-field content within the primary eye's field of view, while keeping the near-field content in the non-primary eye's field of view clear. The blur gradient is 0.1 to 0.25D / step, until the normalized target sharpness V ≤ 0.7. The progressive blurring of the dynamic blur controller is implemented through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V in its progressive blurring is the same as in step S33.
[0061] S43. Restore the near-field content within the dominant eye's field of view to full clarity;
[0062] S5. Competition suppression loop: Controlled by the central control module, steps S3-S4 are executed alternately until the loop is repeated N times.
[0063] In a further preferred technical solution, the central control module controls the alternating execution of steps S3-S4 until the cycle is repeated 10 to 15 times.
[0064] The beneficial effects of this invention are:
[0065] 1. Precise inhibition and regulation
[0066] By using dynamic blur control of the eye-separated vision, the dominant eye's distance vision advantage and the non-dominant eye's near vision advantage are forcibly activated. Clinical data show that the abnormality correction rate reaches 86%, which is significantly improved compared to the traditional occlusion method (62%).
[0067] 2. Contextualized Function Adaptation
[0068] Multi-view distance dynamic scene simulation (5m / 1m / 40cm) fits the daily eye use needs, and combined with voice prompts to strengthen conditioned reflexes, the near vision improvement reaches level 2.1 (Jaeger scale), far exceeding the level 1.4 of traditional training.
[0069] 3. Personalized parameter engine
[0070] The training pace was dynamically adjusted based on age, neuroplasticity, and postoperative recovery data (e.g., a 5-second S-shaped blur curve was used for older users), resulting in a training comfort score of 8.5 points (out of 1-10), which is 60% higher than traditional methods.
[0071] 4. Collaborative Technological Innovation
[0072] The physical split-eye structure (optical lens isolation) works in tandem with the Gaussian blur algorithm (pixel-level sharpness control) to achieve a dual suppression effect of "precise hardware shading + dynamic software adjustment", avoiding the delay error of traditional electronic shading. Attached Figure Description
[0073] Figure 1 This is a schematic diagram of the system structure of the present invention.
[0074] Figure 2 This is a schematic diagram of the training process of this training system.
[0075] Figure 3 This is a schematic diagram of the initial state of a scene image, including both foreground and background content.
[0076] Figure 4 This is a diagram illustrating the blurred state of distant content in the field of vision of a non-dominant eye.
[0077] Figure 5 This is a schematic diagram illustrating the blurred state of near-field content in the dominant eye's field of vision. Detailed Implementation
[0078] The invention will be further described below with reference to the accompanying drawings.
[0079] Example 1:
[0080] like Figure 1 As shown, the present invention describes a VR-based binocular coordination vision training system, which includes the following modules:
[0081] The VR binocular display module physically isolates the display areas of the left and right eye fields of view through its optical lens barrels;
[0082] The dual-eye independent rendering engine supports simultaneous rendering of scene images containing both near and far objects in the display areas of both eyes' visual fields, and independently renders the blurring effects of the near and far objects in the scene images for each eye. The scene image resolution is ≥4K, the near object refers to the image with a viewing distance of 40cm and contains dynamic texture details, and the far object refers to the image with a viewing distance of 5m or more.
[0083] The dynamic blur controller, invoked by the independent rendering engine for each eye, dynamically controls the progressive blurring process and effect of the blurred object. Specifically, it independently adjusts the blur radius of the Gaussian blur shader for the image displayed in the visual field of a single eye according to a set value. The progressive blurring of the dynamic blur controller is achieved through the Gaussian blur shader.
[0084] The user information entry module is used to enter the user's age and set the user's dominant eye.
[0085] The voice prompt module is used to issue voice prompts to the user before the gradual blurring begins, guiding the user to adjust their attention to the blurred object;
[0086] An eye-tracking module is used to monitor eye-tracking latency in real time;
[0087] The central control module is used to read user information, set the initial fuzzy object based on the user information, control the switching of fuzzy objects, control the duration of progressive fuzzing, and control the content and timing of language prompts.
[0088] In this embodiment, the user's dominant eye is defined as the farsighted dominant eye.
[0089] Combination Figure 2 The specific control methods of this system are as follows:
[0090] S1. User Information Entry: The user information entry module collects and enters the user's age and sets the user's dominant eye.
[0091] S2. Device Initialization Settings: The central control module controls the independent rendering engine to generate scene images containing both far and near views within the dual-eye display area, and controls the dynamic blur controller to initialize settings so that both far and near views in the scene images within the dual-eye display area are clear; for example... Figure 3 As shown.
[0092] S3, Long-distance image competition suppression, its blurring effect is as follows: Figure 4 As shown, the specific steps include the following:
[0093] S31. The central control module retrieves the user's primary eye information from the user information input module, and based on the user's primary eye information, sets the distant content in the non-primary eye field of view as the initial blurred object, and transmits the initial blurred object information to the split-eye independent rendering engine.
[0094] S32. The central control module first sends an instruction to the voice prompt module. The voice prompt module responds and sends a first voice instruction to the user, prompting the user to continue to pay attention to the distant content in the scene image.
[0095] S33. The split-eye independent rendering engine calls the dynamic blur controller to progressively blur distant content in the non-primary eye's field of view while keeping distant content in the primary eye's field of view clear. The blur gradient is 0.1D / step until the standardized target sharpness V ≤ 0.7. The duration of progressive blurring is set based on user information, with the following rules: 3 seconds when the user's age is ≤ 55 years old; 5 seconds when the user's age is > 55 years old. During progressive blurring, eye tracking delay is monitored in real time. If the delay is > 500ms, the blurring duration is extended by 0.5 seconds.
[0096] The progressive blurring of the dynamic blur controller is achieved through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V is as follows:
[0097] When V∈[0.7,1.0], R=4×(1.0-V) / 0.3;
[0098] When V∈[0.3,0.7), R=4+6×(0.7-V) / 0.4;
[0099] R: Gaussian blur radius (Blur Radius), in pixels (px);
[0100] V: Visual Acuity, normalized value range 0.3 to 1.0;
[0101] Wherein, target sharpness V is a standardized display sharpness parameter that characterizes the recognizability of target images in a VR scene. V = 1.0 represents a blur-free state, and V = 0.7 represents the minimum recognizable threshold, as detailed below:
[0102] V=1.0: The target is completely clear (no blurring).
[0103] V = 0.7: The target has reached the minimum identifiable threshold (ISO 8596:2017 standard);
[0104] V = 0.3: The target is severely blurred (maximum blur radius 10px);
[0105] "Target sharpness V" refers to the ratio of the smallest recognizable feature size of the target image after Gaussian blurring in a VR display system to the same feature size in the baseline sharp state. The calculation formula is as follows:
[0106] V=\frac{S_{\text{base}}}{S_{\text{blur}}}
[0107] in:
[0108] $S_{\text{base}}$: Pixel size of key features of the target (such as the Landolt C-ring notch) without blurring;
[0109] $S_{\text{blur}}$: The smallest pixel size at which the same feature can be recognized by standard machine vision algorithms after blurring (according to ISO 8596:2017).
[0110] Note: The V value ranges from 0.3 to 1.0 and is precisely controlled by the Shader algorithm. It is unrelated to human visual ability.
[0111] S34. Restore the distant view content in the non-dominant eye's field of vision to full clarity;
[0112] S4. Close-range image competition suppression: Its blurring effect is as follows Figure 5 As shown, the specific steps are as follows:
[0113] S41. The central control module sends a blur object switching command to the split-eye independent rendering engine. The split-eye independent rendering engine responds to the command and switches the blur object to the foreground content within the main eye's field of view.
[0114] S42. The central control module first sends an instruction to the voice prompt module. The voice prompt module responds and sends a second voice instruction to the user, prompting the user to continue to pay attention to the close-up content in the scene image.
[0115] The independent rendering engine for each eye calls the dynamic blur controller to progressively blur the near-field content within the primary eye's field of view, while keeping the near-field content in the non-primary eye's field of view clear. The blur gradient is 0.1D / step, until the normalized target sharpness V ≤ 0.7. The progressive blurring of the dynamic blur controller is implemented through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V in its progressive blurring is the same as in step S33.
[0116] S43. Restore the near-field content within the dominant eye's field of view to full clarity;
[0117] S5. Competition suppression loop: Controlled by the central control module, steps S3-S4 are executed alternately until the loop is repeated 10 times.
[0118] Furthermore, in the specific implementation of this embodiment, the inventor recommends that users set it to 10 times per group, use it twice a day, and continue using it for 2 weeks.
[0119] Example 2:
[0120] Based on Example 1, the setting of the fuzzy gradient in steps S33 and S42 is adjusted to 0.2D / step. The number of iterations of alternating steps S3-S4 in the competition suppression loop of step S5 is adjusted to 12.
[0121] In the specific implementation of this embodiment, the inventor recommends that users set the actual usage cycle as follows: 13 times / set, twice a day, for 3 weeks.
[0122] Example 3:
[0123] Based on Example 1, the fuzzy gradient setting in steps S33 and S42 is adjusted to 0.25D / step. The number of iterations of alternating steps S3-S4 in the competition suppression loop of step S5 is adjusted to 15.
[0124] In the specific implementation of this embodiment, the inventor recommends that users set the actual usage cycle as follows: 15 times / set, once a day, for 4 weeks.
[0125] The following is user experimental data using this training system:
[0126] 1. Experimental Design
[0127] Experimental objective: To verify the effectiveness of the VR-based visual training system (experimental group) in improving binocular inhibition relationship and enhancing near / far visual acuity, and to compare the differences with traditional occlusion training (control group).
[0128] Subject grouping:
[0129] • Experimental group (VR training): 50 patients (aged 45-65 years) still had binocular suppression abnormalities (competition between the dominant eye and the myopic eye) one month after corneal laser presbyopia correction surgery.
[0130] • Control group (traditional occlusion training): 50 matched users, who underwent alternating occlusion training (20 minutes per day).
[0131] Training cycle: 4 weeks, 5 days a week, 20 minutes a day.
[0132] 2. Evaluation Indicators
[0133] (1) Vision:
[0134] Distance visual acuity (5m LogMAR visual acuity chart)
[0135] Near vision (40cm Jaeger near vision chart)
[0136] (2) Binocular suppression:
[0137] The inhibition status (normal / abnormal) was assessed using the Worth 4-point lamp test;
[0138] Suppression ratio (the proportion of myopia suppressed by the dominant eye, measured by a vision function instrument).
[0139] (3) Subjective comfort:
[0140] User questionnaire (1-10 points, 10 points is the best).
[0141] 3. Experimental Results
[0142] index Experimental group (VR training) Control group (traditional training) P-value (significance) LogMAR (LogMAR) - Improves Distance Vision 0.12±0.05 0.08±0.06 P<0.01 Near vision improvement (Jaeger) 2.1±0.3 1.4±0.4 P<0.001 Suppressing abnormal correction rate 86％ 62％ P<0.05 Training comfort level (1-10 points) 8.5±1.2 5.3±1.8 P<0.001
[0143] Key conclusions:
[0144] (1) More significant improvement in vision: The VR training group showed significantly greater improvement in both far and near vision than the traditional training group (P<0.01), especially in near vision (Jaeger improved by 2.1 levels vs. 1.4 levels).
[0145] (2) More efficient inhibition relationship correction: 86% of VR training users achieved normal inhibition (dominant eye dominates distance vision, myopic eye dominates near vision) 4 weeks after the operation, while only 62% of traditional training users achieved inhibition.
[0146] (3) Higher user compliance: The comfort score of VR training (8.5 points) is significantly better than that of traditional occlusion training (5.3 points), and it is easier to persist due to the interactive and scenario-based design.
[0147] 4. Typical User Analysis
[0148] User 1 (Experimental Group):
[0149] • Preoperative: The dominant eye (right eye) was fully corrected, while the left eye retained -1.5D myopia. Postoperatively, the left eye failed to suppress distance vision (double vision).
[0150] • After VR training:
[0151] Week 1: Single-eye training phase, the dominant eye's distance vision clarity improves by 30%;
[0152] Week 2: Near vision clarity improved by 40% in nearsighted individuals;
[0153] Week 4: Worth 4-dot test showed normal inhibition, with both distance and near vision reaching 20 / 20.
[0154] User 2 (control group):
[0155] • Under the same baseline conditions, after 4 weeks of traditional occlusion training, there was still partial suppression in the left eye when looking at distant objects (Worth 4-dot light abnormality), so the training period needs to be extended.
[0156] 5. Advantages compared to traditional solutions
[0157] Comparison Dimensions VR Training Solution Traditional occlusion training accuracy Dynamic control of split eyes precisely enhances the advantage of the target eye. Relies on user's subjective cooperation, resulting in large errors. Scene adaptability Simulate real-world near / far scenes to enhance conditioned reflexes. Static training cannot simulate dynamic switching. Data feedback Real-time recording of clarity processing time and optimization of parameters Relying on manual observation, without quantitative data User acceptance High (gamified interaction) Low (high repeatability, easy to tire)
[0158] The above shows and describes the basic principles, main features, and advantages of this solution.
[0159] Point. Those skilled in the art should understand that this solution is not limited to the above embodiments.
[0160] The embodiments and descriptions in the specification are merely illustrative of the principles of this solution and should not be construed without departing from this solution.
[0161] Within the framework of the spirit and scope, this plan will undergo various changes and improvements.
[0162] All of these fall within the scope of this claim. The scope of protection claimed by this claim is defined by the attached rights.
[0163] The definition of the claim and its equivalent.
Claims
1. A VR-based binocular coordination vision training system, characterized in that, include: The VR binocular display module physically isolates the display areas of the left and right eye fields of view through its optical lens barrels; The dual-eye independent rendering engine supports simultaneous rendering of scene images containing both distant and near-field content in the display area of both eyes' field of vision, and independently renders the blurring effect of distant and near-field content in the scene images of both eyes. The dynamic blur controller, invoked by the eye-independent rendering engine, dynamically controls the progressive blurring process and effect of the blurred object. Specifically, it independently adjusts the Gaussian blur shader radius of the image displayed in the visual field of a single eye according to a set value. The progressive blurring of the dynamic blur controller is achieved through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V is as follows: When V∈[0.7,1.0], R=4×(1.0-V) / 0.3; When V∈[0.3,0.7), R=4+6×(0.7-V) / 0.4; R: Gaussian blur radius, in pixels; V: Target sharpness, normalized value range 0.3 to 1.0; Among them, the target sharpness V is a standardized display sharpness parameter that characterizes the recognizability of the target image in the VR scene. V = 1.0 indicates a no-blur state, and V = 0.7 indicates the minimum recognizable threshold. The blur gradient of the dynamic blur controller is 0.1 to 0.25D / step, and blurring stops when the standardized target sharpness V ≤ 0.7; The user information entry module is used to enter the user's age and set the user's dominant eye. The voice prompt module is used to issue voice prompts to the user before the gradual blurring begins, guiding the user to adjust their attention to the blurred object; An eye-tracking module is used to monitor eye-tracking latency in real time; The central control module is used to read user information, set the initial fuzzy object based on the user information, control the switching of fuzzy objects, control the duration of progressive fuzzing, and control the content and timing of language prompts.
2. The VR-based binocular coordination vision training system as described in claim 1, characterized in that, The scene image has a resolution of ≥4K, and its near-field content refers to an image with a viewing distance of 40cm that includes dynamic texture details; its far-field content refers to an image with a viewing distance of 5m or more.
3. The VR-based binocular coordination vision training system as described in claim 1, characterized in that, The central control module retrieves the user's dominant eye information from the user information input module and sets the initial fuzzy object of the dynamic fuzzy controller based on the user's dominant eye information.
4. The VR-based binocular coordination vision training system as described in claim 1, characterized in that, The central control module retrieves user age information from the user information input module to determine the duration of progressive fuzzification of the dynamic fuzzy controller. The determination rule is as follows: when the user's age is ≤55 years old, it is set to 3 seconds; when the user's age is >55 years old, it is set to 5 seconds.
5. The VR-based binocular coordination vision training system as described in claim 1, characterized in that, During the progressive blurring process of the dynamic blur controller, the eye-tracking module monitors the eye-tracking delay in real time. If the delay is greater than 500ms, it sends a signal to the central control module, which then controls the dynamic blur controller to extend the blurring duration by 0.5 seconds.
6. The VR-based binocular coordination vision training system as described in claim 1, characterized in that, The central control module controls the split-eye independent rendering engine to alternately and cyclically render the progressive blurring effect of far and near content in the display area of the VR dual-eye display module, and the number of cycles is controlled to be between 10 and 15.
7. The VR-based binocular coordination vision training system as described in claim 1, characterized in that, The user's dominant eye is defined as the farsighted visual dominant eye.
8. A control method for the binocular coordination vision training system as described in any one of claims 1-7, characterized in that, The control method includes the following steps: S1. User Information Entry: The user information entry module collects and enters the user's age and sets the user's dominant eye. S2. Device initialization settings: The central control module controls the eye-separated independent rendering engine to generate scene images containing far and near content in the display areas of both eyes, and controls the dynamic blur controller to initialize settings so that the far and near content of the scene images in the display areas of both eyes is clear. S3. Long-range image competition suppression: This includes the following steps: S31. The central control module retrieves the user's primary eye information from the user information input module, and based on the user's primary eye information, sets the distant content in the non-primary eye field of view as the initial blurred object, and transmits the initial blurred object information to the split-eye independent rendering engine. S32. The central control module first sends instruction information to the voice prompt module, the voice prompt module responds and sends a first voice instruction to the user, prompting the user to continue to pay attention to the distant content in the scene image; S33. The independent rendering engine for each eye calls the dynamic blur controller to progressively blur distant content within the non-primary eye's field of vision, while maintaining clarity of distant content within the primary eye's field of vision. The blur gradient is 0.1–0.25D / step, until the standardized target sharpness V ≤ 0.
7. The duration of the progressive blurring is set based on user information, with the following rules: 3 seconds for users aged ≤ 55, and 5 seconds for users aged > 55. During the progressive blurring process, eye-tracking delay is monitored in real time; if the delay > 500ms, the blurring duration is extended by 0.5 seconds. The progressive blurring of the dynamic blur controller is achieved through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V is as follows: When V∈[0.7,1.0], R=4×(1.0-V) / 0.3; When V∈[0.3,0.7), R=4+6×(0.7-V) / 0.4; R: Gaussian blur radius, in pixels; V: Target sharpness, normalized value range 0.3 to 1.0; Among them, the target sharpness V is a standardized display sharpness parameter that characterizes the recognizability of the target image in the VR scene. V = 1.0 indicates a no-blur state, and V = 0.7 indicates the minimum recognizable threshold. S34. Restore the distant view content in the non-dominant eye's field of vision to full clarity; S4. Close-range image competition suppression: The specific steps are as follows: S41. The central control module sends a blur object switching command to the split-eye independent rendering engine. The split-eye independent rendering engine responds to the command and switches the blur object to the near-field content within the main eye's field of view. S42. The central control module first sends an instruction message to the voice prompt module, the voice prompt module responds and sends a second voice instruction to the user, prompting the user to continue to pay attention to the close-up content in the scene image; The independent rendering engine for each eye calls the dynamic blur controller to progressively blur the near-field content within the primary eye's field of view, while keeping the near-field content in the non-primary eye's field of view clear. The blur gradient is 0.1 to 0.25D / step, until the normalized target sharpness V ≤ 0.
7. The progressive blurring of the dynamic blur controller is implemented through a Gaussian blur shader, and the mapping relationship between the blur radius R and the target sharpness V in its progressive blurring is the same as in step S33. S43. Restore the near-field content within the dominant eye's field of view to full clarity; S5. Competition suppression loop: Controlled by the central control module, steps S3-S4 are executed alternately until the loop is repeated N times.
9. The control method as described in claim 8, characterized in that, The central control module controls the alternating execution of steps S3-S4 until the cycle is repeated 10 to 15 times.