A method and device for refractive compensation based on dual-mode perception

Abstract

This invention relates to the field of refractive power adjustment technology and discloses a refractive compensation method and device based on bimodal perception. The method includes first establishing a coordinate system of the head-mounted device, defining a first region and a second region, allowing the user to switch gaze back and forth between the two regions, and collecting bimodal perception data including physiological signal modality and spatial motion modality; then, based on the bimodal perception data, determining whether an emergency or stable accommodation event occurs. The former triggers an advanced compensation acceleration mechanism, while the latter uses a conventional compensation mechanism. This method solves the problem of asynchronous traditional compensation algorithms and physiological responses through bimodal data fusion and a hierarchical compensation strategy. In high-frequency gaze switching scenarios, it can effectively reduce ineffective accommodation movements, alleviate ciliary muscle fatigue, and improve the accuracy and physiological adaptability of refractive compensation.

Description

Technical Field

[0001] This invention relates to the field of refractive power adjustment technology, and more specifically, to a refractive compensation method and apparatus based on bimodal sensing. Background Technology

[0002] In the prior art, Chinese patent application CN118131475A discloses a head-mounted display device and methods for refractive power detection, interpupillary distance compensation, and gaze accommodation. The disclosed head-mounted display device obtains the refractive power by detecting the markings on the refractive correction lenses, thus achieving interpupillary distance compensation and gaze accommodation. Chinese patent CN115736814B, in its authorization announcement, proposes a device, lens, method, and storage medium for fitting the uncorrected retinal refractive power. This prior art method for fitting the uncorrected retinal refractive power achieves personalized refractive correction through rotational symmetry plane fitting.

[0003] However, with the increasing demand for high-frequency fixation switching in educational and daily life scenarios, frequent fixation switching between near and far distances can lead to ciliary muscle accommodation lag. For example, frequent fixation switching between a student's desk and the blackboard can cause ciliary muscle accommodation lag. Failure to provide timely refractive compensation can easily lead to high-frequency accommodation fatigue of the user's ciliary muscle. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of existing technologies, this invention provides a refractive compensation method and device based on bimodal perception. By establishing a coordinate system for the head-mounted device, it simultaneously acquires bimodal data such as ciliary muscle electromyography (CME) signals, head posture, and gaze distance. Based on multi-dimensional features including distance change rate, head motion gradient, trajectory curvature, and CME slope, it accurately determines emergency and stable accommodation events. For emergency scenarios, it dynamically shortens accommodation delay and pre-calculates the target refractive power; for stable scenarios, it employs baseline correction and periodic compensation, forming a hierarchical decision-making mechanism. This solution achieves real-time synchronization between lens refractive compensation and ciliary muscle accommodation rhythm, effectively reducing ineffective accommodation movements during high-frequency switching, minimizing abnormal fluctuations in refractive power, significantly improving physiological adaptability and compensation accuracy in dynamic visual scenarios, alleviating ciliary muscle fatigue at its source, and providing an innovative solution for myopia prevention and control.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A refractive compensation method based on bimodal sensing includes:

[0007] Establish a coordinate system O-XYZ for the head-mounted device, and define a first region and a second region in the coordinate system O-XYZ; the user performs a back-and-forth gaze switching between the first region and the second region at different switching speeds, and collect bimodal perception data during the user's gaze switching process;

[0008] Based on bimodal perception data during user gaze switching, emergency adjustment events and stable adjustment events are determined. If an emergency adjustment event is determined, an advance compensation acceleration mechanism is triggered; if a stable adjustment event is determined, a conventional compensation mechanism is used.

[0009] Furthermore, the method for establishing the O-XYZ coordinate system of the head-mounted device is as follows: the origin O is the midpoint of the line connecting the pupils of the user's two eyes, the X-axis runs from left to right along the line connecting the pupils of the two eyes, the Y-axis points to the top of the head, and the Z-axis is defined as the direction of the line of sight directly in front.

[0010] Furthermore, the method for defining the first and second regions in the coordinate system O-XYZ is as follows: the interval Z∈[a,b] is defined as the first region, and the interval Z∈[c,d] is defined as the second region, where a,b,c,d are positive real numbers and a<b<c<d.

[0011] Furthermore, the dual-modality refers to the physiological signal mode and the spatial motion mode, and the dual-modal sensing data includes physiological signal data and spatial motion data;

[0012] The physiological signal data refers to the electrociliary muscle signal;

[0013] The spatial motion data includes the user's head posture data, gaze distance, and three-dimensional gaze point coordinates.

[0014] Furthermore, the method for determining emergency adjustment events and stable adjustment events is as follows:

[0015] Based on head pose data and gaze distance, the distance change rate V is calculated. D Head motion gradient G H And the three-dimensional trajectory curvature κ; calculate the slope ke of the change in the ciliary muscle electromyography signal;

[0016] Based on the distance change rate V D Head motion gradient G H The three-dimensional trajectory curvature κ and the slope ke of the ciliary electromyography signal are used to determine emergency and stable regulatory events.

[0017] Furthermore, the calculated distance change rate V D The method is as follows: using a sliding time window, the distance change rate V is calculated based on the gaze distance. D ;

[0018] Furthermore, the curvature of the three-dimensional trajectory The calculation method is as follows:

[0019] A global space curve is generated from the coordinates of the 3D gaze point; the current gaze point is selected from the global space curve. Compared to the past One gaze point The curvature of the constructed local space curve is calculated as the curvature of the three-dimensional trajectory. .

[0020] Furthermore, the distance change rate V D Head motion gradient G H The method for determining emergency and stable regulatory events is based on the three-dimensional trajectory curvature κ and the slope ke of the ciliary electromyography signal.

[0021] like and and and >k th If the event is positive, it is classified as an emergency adjustment event; otherwise, it is classified as a stable adjustment event. The preset distance change rate threshold, The preset head motion gradient threshold, k is the preset trajectory curvature threshold. th This is a preset threshold for the slope of change.

[0022] Furthermore, the advanced compensation acceleration mechanism is as follows: dynamically shortening the adjustment delay parameter. The shortened compensation delay time is obtained. ,based on Calculate the target refractive power in advance ,according to Adjust the lens; the adjustment delay parameter It was acquired synchronously during the collection of bimodal perception data during the user's gaze switching process.

[0023] A refractive compensation device based on bimodal sensing, used to implement the aforementioned refractive compensation method based on bimodal sensing, the device comprising:

[0024] Partitioning module: Used to establish the coordinate system O-XYZ of the head-mounted device, and to define the first and second regions in the coordinate system O-XYZ;

[0025] Dual-modal data acquisition module: The user performs back-and-forth gaze switching between the first and second regions at different switching speeds, and the module collects dual-modal perception data during the user's gaze switching process; the dual-modality refers to physiological signal mode and spatial motion mode;

[0026] Adjustment event determination module: Based on bimodal perception data during user gaze switching, it determines emergency adjustment events and stable adjustment events;

[0027] Refractive compensation module: If an emergency accommodation event is identified, an advanced compensation acceleration mechanism is triggered; if a stable accommodation event is identified, a conventional compensation mechanism is used.

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0029] This invention achieves precise positioning of the user's gaze space by constructing a coordinate system for the head-mounted device and dividing it into a first region and a second region. Combined with dual-modal sensing data acquisition, it simultaneously acquires ciliary muscle electromyography (CME) signals and spatial motion data such as head posture and gaze distance, characterizing the accommodation scenario from both physiological response and behavioral features. Based on multi-dimensional feature calculations of distance change rate, head movement gradient, three-dimensional trajectory curvature, and ciliary muscle CME signal change slope, it accurately determines emergency and stable accommodation events, enabling compensation strategies to dynamically adjust according to accommodative load. For emergency accommodation events, the proactive compensation acceleration mechanism dynamically shortens the accommodation delay parameter and combines it with predicted accommodation lag to adjust lens refractive power in advance, offsetting the physiological delay of the ciliary muscle. The conventional compensation mechanism, on the other hand, periodically calculates refractive compensation and performs baseline correction, reducing computational load while maintaining accuracy. The combination of these two approaches forms a hierarchical decision-making mechanism that covers the entire range of gaze switching scenarios, enabling real-time synchronization between lens refractive compensation and ciliary muscle accommodation rhythm. This effectively reduces ineffective accommodation movements caused by delays, lowers abnormal fluctuations in refractive power, alleviates high-frequency accommodation fatigue of the ciliary muscle, and significantly improves the accuracy of refractive compensation and human physiological adaptability in dynamic visual scenarios. It fundamentally solves the core problem of the disconnect between compensation strategies and real physiological responses in traditional solutions.

[0030] The technical solution of this invention can be widely adapted to the dynamic refractive compensation needs in the field of light field display (such as holographic display, naked-eye 3D, stereoscopic display, etc.). In light field display scenarios, the user's line of sight often switches frequently with changes in the depth of the virtual / real scene. This invention, through dual-modal data fusion and layered compensation strategies, can accurately match the dynamic refractive requirements of light field display devices, effectively improving the visual clarity and physiological comfort when virtual objects are integrated with the real environment, and providing key technical support for optimizing the human-computer interaction experience of light field display devices. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0032] Figure 1 This is a flowchart illustrating the principle of a refractive compensation method based on bimodal sensing in this invention.

[0033] Figure 2 This is a flowchart of a method for determining emergency accommodation events and stable accommodation events in a refractive compensation method based on bimodal sensing according to the present invention.

[0034] Figure 3 This is a schematic diagram illustrating a scenario of rapidly changing gaze distance in an embodiment of the present invention;

[0035] Figure 4 This is a schematic diagram illustrating a scenario where the gaze distance changes slowly in an embodiment of the present invention;

[0036] Figure 5 This is a flowchart of a method for obtaining an individual's initial compensation coefficient matrix A in a bimodal sensing-based refractive compensation method according to the present invention.

[0037] Figure 6 This is a functional block diagram of a refractive compensation device based on dual-modal sensing in this invention.

[0038] Explanation of the attached diagram labels: 1. User, 2. Desk, 3. Blackboard. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] Example 1

[0041] Please see Figure 1 As shown, this embodiment provides a refractive compensation method based on bimodal sensing, including:

[0042] Step S1000: Establish the head-mounted device body coordinate system O-XYZ, and define a first region and a second region in the coordinate system O-XYZ; the user performs back-and-forth gaze switching between the first region and the second region at different switching speeds, and collects bimodal perception data during the user's gaze switching process.

[0043] Further, step S1000 includes:

[0044] Step S1100: Establish the head-mounted device body coordinate system O-XYZ, and define the first region and the second region in the coordinate system O-XYZ;

[0045] Further, step S1100 includes:

[0046] Step S1110: Establish the head-mounted device body coordinate system O-XYZ with the midpoint of the line connecting the user's pupils as the origin O. This coordinate system is a right-handed Cartesian coordinate system. The X-axis runs from left to right along the line connecting the pupils, the Y-axis points to the top of the head, and the Z-axis is defined as the direction of the line of sight.

[0047] Step S1120: Define a first region and a second region in the head-mounted device body coordinate system O-XYZ; the near distance interval Z∈[a,b] is defined as the first region, and the far distance interval Z∈[c,d] is defined as the second region, where a,b,c,d are positive real numbers and a<b<c<d.

[0048] Specifically, the X-axis runs along the line connecting the pupils of both eyes, from left to right, and is parallel to the plane of the first region (such as a desk), representing the horizontal movement of the user's head. The Y-axis is perpendicular to the line connecting the pupils, pointing towards the top of the head, representing the vertical movement of the head. The Z-axis runs along the line directly in front of the user's line of sight, perpendicular to the lens plane, representing the distance to the target being gazed upon (i.e., the depth of field). This coordinate system is calibrated in real time using a nine-axis inertial sensor (integrated into the bridge of the glasses frame) to ensure that the coordinate system dynamically adjusts with head posture and remains consistent with the user's visual axis. The midpoint of the pupil is chosen as the origin because the pupil position directly reflects the geometric center of the line of sight, providing the most accurate mapping of the spatial position of the target being gazed upon. The orthogonal design of the X / Y / Z axes conforms to ergonomics, decomposing head movement into three independent dimensions: horizontal, vertical, and depth of field, facilitating subsequent quantitative analysis of the gaze distance (Z-axis coordinates) and head rotation (X / Y-axis posture angles). Using a non-pupil origin (such as the device's geometric center) could lead to deviations between the line of sight coordinates and the actual gaze point, affecting the accuracy of distance measurements.

[0049] In the O-XYZ coordinate system, two types of core functional areas are divided according to the Z-axis coordinate interval:

[0050] The first region is defined as the Z-axis interval [a, b]. For example, the value is [0.2m, 0.6m], which corresponds to the typical distance range between a primary school student's desk and their eyes (covering near-field objects such as books and stationery).

[0051] The second region is defined as the Z-axis interval [c, d]. For example, the value is [2m, 15m], which covers the common distance range of classroom blackboards (adapting to different classroom space sizes).

[0052] The interval parameters a, b, c, and d satisfy a < b < c < d. Environmental depth data is measured in real-time using the glasses' built-in Time-of-Flight (ToF) sensor, and dynamically calibrated based on historical statistical data to ensure the region division conforms to actual usage scenarios. The Z-axis is used as the primary dividing line because the ciliary muscle's accommodative load is mainly caused by changes in gaze distance (the lens curvature needs to be adjusted when switching between near and far distances). The clear definition of the first and second regions simplifies the complex three-dimensional space into interval judgments along the depth direction, significantly reducing the computational complexity of subsequent algorithms. Without region division, the system would need to process gaze points across the entire spatial range in real-time, leading to wasted computational resources and decreased efficiency in recognizing accommodative events.

[0053] Step S1100 addresses the core issue of "lack of spatial positioning reference" by constructing a body coordinate system and dividing functional areas, laying the geometric foundation for subsequent data acquisition and accommodation event analysis. By incorporating user head posture, gaze point coordinates, and lens refractive power adjustment direction into the same coordinate system, physiological signals (such as ciliary electromyography signals) and spatial motion data (such as changes in gaze distance) can be spatiotemporally aligned. For example, when a gaze point is detected rapidly moving from the first area (Z=0.3m) to the second area (Z=5m), the distance change ΔD=4.7m provided by the coordinate system can be directly used as a quantitative indicator of accommodation load. Through Z-axis interval division, near (desk) and far (blackboard) gaze targets are clearly distinguished, specifically capturing the most frequent accommodation switching scenarios in elementary school classrooms (according to educational scenario statistics, the average number of switching times per class period reaches 80-120). Without area definition, the system cannot quickly identify effective accommodation events, leading to invalid data interfering with the decision-making of compensation strategies. The real-time calibration function of the coordinate system (based on inertial sensors) ensures that head movements do not affect the accuracy of region division. Even when students look down to write (change in posture on the Y-axis) or turn their heads to communicate (rotation on the X-axis), the depth measurement on the Z-axis can still accurately reflect the actual distance of the gaze target. This robust design avoids the positioning errors of traditional fixed coordinate systems when head posture changes (such as misjudging the first region as a farther distance when looking down).

[0054] Step S1200: Based on the head-mounted device body coordinate system O-XYZ, the user performs back-and-forth gaze switching between the first and second regions at different switching speeds, and collects bimodal perception data during the user's gaze switching process.

[0055] The dual-modality refers to the physiological signal modality and the spatial motion modality. The dual-modal sensing data includes physiological signal data and spatial motion data. The physiological signal data refers to ciliary electromyography (CEMG) signals. The spatial motion data includes the user's head posture data, gaze distance, and three-dimensional gaze point coordinates. A global spatial curve is generated from the three-dimensional gaze point coordinates.

[0056] The method for collecting the user's head posture data is as follows: a nine-axis inertial sensor is integrated into the bridge of the nose of the glasses frame to collect head posture data; the head posture data includes the head pitch angle. Yaw angle Roll angle ,as well as , and rate of change ;

[0057] The method for acquiring the ciliary muscle electromyography (CEMG) signal is as follows: an n1-channel dry electrode bioelectric sensor is arranged on the inner side of the temple of the spectacle to acquire the CEMG signal.

[0058] Specifically, step S1200 integrates a multi-sensor module to simultaneously acquire spatial motion data reflecting gaze behavior and electrical signal data reflecting the physiological response of the ciliary muscle, forming a dual-modal perception dataset. In this embodiment, the physiological signal modality specifically refers to the ciliary muscle electromyography (EMG) signal, which is acquired by an n1-channel dry electrode sensor located on the inner side of the temple. The dry electrode is made of silver / silver chloride material, which can adhere to the periorbital skin without conductive gel, resulting in high comfort during long-term wear. After the signal is bandpass filtered from 20-500Hz (to remove low-frequency noise and high-frequency interference), the root mean square (RMS) value is extracted as a quantitative indicator of muscle activation.

[0059] Spatial motion modality: includes head posture data, gaze distance, and 3D gaze point coordinates. Head posture data is acquired via a nine-axis inertial sensor (fusion of accelerometer, gyroscope, and magnetometer) located at the bridge of the nose frame, outputting the head pitch angle θ (rotation angle around the X-axis), yaw angle ϕ (rotation angle around the Y-axis), roll angle ψ (rotation angle around the Z-axis), and the real-time rate of change of each angle. The gaze distance reflects the direction and intensity of head movements. The gaze distance is measured in real time by a ToF sensor on the lens, indicating the distance to the target in the direction of the gaze. The three-dimensional gaze coordinates are combined with the eye-tracking module (an infrared camera integrated above the lens) and the body coordinate system to map the retinal imaging position to the O-XYZ coordinate system, generating a continuous gaze sequence. This sequence is further fitted into a global spatial curve for analyzing the curvature and directional changes of the gaze trajectory.

[0060] Guide the user to perform N1 round-trip gaze switching operations (from the first region to the second region and back). Each switch includes the following key steps:

[0061] Triggering condition: When the gaze point stays in the first region for ≥100ms, and then enters the second region for the first time and stays there for ≥100ms, it is considered a valid switch.

[0062] Data storage: Each time the corresponding bimodal perception data (including θ, ϕ, ψ and their rate of change, gaze distance sequence, electromyography RMS sequence, and three-dimensional gaze point coordinates) is switched, it is stored synchronously with timestamps to form a dataset containing N1 samples, which is used for subsequent model training and event determination.

[0063] By simultaneously collecting changes in fixation distance (spatial motion data) and peak delays in electromyographic signals (physiological data), a precise characterization of an individual's "accommodative lag characteristics" can be achieved. Combining head posture data (θ, ϕ, ψ and their rates of change) with fixation point coordinates allows for the differentiation between "active head-turning fixation" and "eye-rotation fixation." For example, when the head yaw angle ϕ changes rapidly, even if the fixation distance remains constant, it may be accompanied by synergistic movement of the ciliary muscle (due to head movement causing minor adjustments in the line of sight), which needs to be included in the accommodative load calculation. If only fixation distance is collected while head movement is ignored, such implicit accommodative needs will be missed, leading to insufficient compensation.

[0064] The establishment of a coordinate system provides a spatial positioning benchmark for data acquisition, while the spatiotemporal characteristics of the dual-modal data feed back into the dynamic calibration of the coordinate system, forming a closed-loop support. When head posture data (θ, ϕ, ψ) shows that the user's head is tilted down by 15°, the coordinate system automatically adjusts the Z-axis direction (tilting with the line of sight) to ensure that the Z-axis interval of the first region is still based on the actual line of sight distance, rather than a fixed vertical distance, avoiding misjudgment of the region due to changes in head posture. Through the correlation analysis between the gaze point coordinates and electromyographic signals, it is possible to identify whether a specific spatial region (such as the upper left corner of the first region) triggers abnormal accommodative load, providing a basis for subsequent personalized region compensation. This collaboration can maintain accurate perception of accommodative needs even in complex classroom environments (such as students tilting their heads to write or looking up at the projector, etc.), fundamentally solving the shortcomings of traditional solutions that "only rely on fixed distance thresholds and ignore individual posture differences," and improving the spatiotemporal accuracy of dynamic refractive compensation to the same frequency level as physiological signal response.

[0065] Step S2000: Based on the bimodal perception data during the user's gaze switching process, determine whether it is an emergency adjustment event or a stable adjustment event; if it is determined to be an emergency adjustment event, trigger the advance compensation acceleration mechanism; if it is determined to be a stable adjustment event, adopt the conventional compensation mechanism.

[0066] Further, step S2000 includes:

[0067] Step S2100: Based on the bimodal perception data during the user's gaze switching process, determine whether it is an emergency adjustment event or a stable adjustment event;

[0068] Furthermore, such as Figure 2 As shown, step S2100 includes:

[0069] Step S2110: Calculate the distance change rate V based on head pose data and gaze distance. D Head motion gradient G H and the curvature κ of the three-dimensional trajectory;

[0070] Further, step S2110 includes:

[0071] Step S2111: Using a sliding time window, calculate the distance change rate V based on the gaze distance. D ;

[0072] Specifically, a sliding time window is used for calculation, with a time window length of [missing information]. Step size is The calculation formula is: in, For the current moment gaze distance, For the first The gaze distance at each sampling time. The sampling time interval, Rate of change of distance This reflects the speed at which the user switches their gaze between near (region 1) and far (region 2) targets. A larger value indicates a faster switching speed. For example, such as... Figures 3-4 As shown, Figure 3 In the middle, the user is in the first At each sampling moment, the eyes are focused on the desk, that is, the first area at close range, in the previous moment. At this time, the user looks at the blackboard, that is, the target is the second area at a distance; the calculated rate of distance change at this time is V. D1 . Figure 4 In the middle, the user is in the first At each sampling moment, the eyes are focused on the desk, and the first area of ​​focus is at close range. At this time, the user's gaze target is still the first region at close range, and the calculated rate of distance change is V. D2 ; Figure 3 V in application scenarios D1 Larger values, due to the delay in the physiological response of the ciliary muscle to complete adjustment, may lead to a deviation between the instantaneous refractive state and the actual requirement. The use of a sliding time window avoids noise interference from single-moment data, and through continuous... Trend analysis at each sampling point accurately captures the transient characteristics of distance changes. Directly using the difference between adjacent time points to calculate speed is susceptible to sensor noise, leading to misjudgments. However, fixed-window averaging can smooth high-frequency fluctuations and preserve the true trend.

[0073] Step S2112, based on the head pose data Calculate the head motion gradient G H ;

[0074] Specifically, This converts the motion velocities in the three directions into a combined gradient value. (Head motion gradient) This comprehensively reflects the intensity of head movement in three directions. A larger gradient indicates more intense head movement. The head movement gradient reflects the synergistic load of ciliary muscle accommodation—when the head rotates rapidly (such as when suddenly looking up at a blackboard accompanied by neck movement), the extraocular muscles need to adjust the direction of vision synchronously, increasing the accommodative pressure on the ciliary muscle. Traditional approaches only focus on distance changes, ignoring such synergistic movements, resulting in compensation strategies that do not fully match the actual physiological load. The introduction of [something] filled this gap.

[0075] Step S2113: Select the current gaze point from the global spatial curve. Compared to the past One gaze point The curvature of the constructed local space curve is calculated as the curvature of the three-dimensional trajectory. ;

[0076] Specifically, ,in, The parametric equation for the fixation point. and These are the first derivative (tangent vector) and the second derivative (curvature vector), respectively.

[0077] Three-dimensional trajectory curvature It reflects the curvature of the gaze trajectory. The larger the value, the more curved the trajectory, meaning the more drastic the shift in gaze. The calculation is based on continuous (For example, M=5) three-dimensional gaze coordinates rely on the real-time localization of the gaze target by the ToF sensor and the coordinate mapping of the eye-tracking module to ensure the spatiotemporal alignment of continuous gaze points. If only a single gaze coordinate is used, it is impossible to capture the dynamic changes of the trajectory. Curvature calculation, through differential geometry methods, effectively identifies "non-linear" gaze shifts. Such scenarios are often accompanied by sudden adjustment needs of the ciliary muscle.

[0078] Step S2120: Calculate the slope ke of the change in the ciliary muscle electromyography signal;

[0079] Specifically, after acquiring ciliary electromyography signals through the n1 channel dry electrode sensor on the inner side of the temple, bandpass filtering (e.g., 5-50Hz) is first performed to remove power frequency noise and high frequency interference, in order to filter out power frequency noise (e.g., 50Hz AC interference) and low frequency drift (e.g., skin electrode contact noise). Then, the root mean square value (RMS) of the signal is calculated using the sliding window method to form a time sequence E(t) characterizing the degree of muscle activation.

[0080] For example, the slope ke is calculated as follows: at the event trigger time T when the gaze distance change is detected... start Then, select T. start To T start The analysis interval is defined as a time window of +300ms. First, the signal baseline value E is determined. baseline Take T start The average value of E(t) within the first 100ms was used as a reference value when the ciliary muscle was not activated. Then, the extreme value of the signal E was searched within the window. peak and its occurrence time T peak (i.e., the maximum or minimum point of E(t)). The slope ke is defined as the linear rate of change of the signal from the baseline value to the extreme value. For example, when a student suddenly looks up at the blackboard, the ciliary muscle needs to relax quickly. The absolute value of the slope of the decrease in the RMS value of the electromyographic signal at this time can also be used as a criterion to ensure the assessment of the responsiveness to the relaxation process. This slope reflects the response speed of the ciliary muscle during the adjustment process. If the signal does not show a valid extreme value within the window, it is judged as invalid data and not included in the subsequent judgment.

[0081] Steps S2110 and S2120, through multi-dimensional feature extraction, solve the core problem of relying on a single criterion for determining adjustment events. , , Ke describes the adjustment scenario from four dimensions: "distance change", "head movement", "eye trajectory" and "physiological response", forming a three-dimensional judgment model. The rate at which distance changes is quantified is the core driving force that triggers the need for adjustment; It captures the synergistic load brought about by head movements, solving the problem that traditional solutions ignore the influence of neck movements; κ It identifies the dynamic characteristics of gaze trajectories, covering high-frequency scenarios of non-linear gaze shifts (such as intensive accommodation when scanning text). Ke verifies the accommodation state at an intrinsic physiological level, ensuring that the judgment results accurately reflect individual muscle response capabilities. and κ The combination of these factors can help predict and adjust the intensity of the load: for example, high... Combining high κ This indicates that significant and frequent lens accommodation is required in a short period of time, with an extremely high risk of hysteresis; through The association with ke can assess the efficiency of physiological regulation: high But low This indicates that head movements exacerbate muscle response delays, requiring stronger proactive compensation.

[0082] Step S2130, based on the distance change rate V D Head motion gradient G H The three-dimensional trajectory curvature κ and the slope ke of the ciliary electromyography signal are used to determine emergency and stable regulatory events.

[0083] like and and and >k th If the event is positive, it is classified as an emergency adjustment event; otherwise, it is classified as a stable adjustment event. The preset distance change rate threshold is used to determine whether the distance change rate meets the "fast" standard. The preset head motion gradient threshold is used to determine whether the head motion is "significant". The preset trajectory curvature threshold is used to determine whether the trajectory has a "sharp turn," k th This is a preset threshold for the slope of change.

[0084] Specifically, step S2120 establishes a multi-feature joint judgment model to achieve accurate classification of emergency and stable regulatory events. The setting of the three thresholds (V1, G1, K1) follows the principle of "physiological response boundary constraints + scenario feature statistics + individual difference adaptation".

[0085] Distance change rate threshold The design is based on the physiological delay characteristics of ciliary muscle accommodation. When the rate of change in fixation distance exceeds the maximum response capacity of the ciliary muscle, instantaneous refractive errors will accumulate significantly. By collecting the peak delay time of ciliary muscle electroencephalogram (EEG) signals from different users during fixation switching, a critical value for the rate of change in distance is determined—this value corresponds to the boundary condition that "the ciliary muscle cannot complete accommodation within the effective time, resulting in lens compensation demand significantly exceeding the physiological response." For example, in a classroom setting, students move from the first area... When quickly looking towards the second region (3m), if the switching time is shorter than the average ciliary muscle accommodation time (approximately 150ms), then the rate of distance change... Taking into account individual differences and sensor noise, the actual threshold Set it slightly below the theoretical value to ensure coverage. The above are examples of quick adjustment scenarios.

[0086] The head motion gradient reflects the combined intensity of head movements in the pitch, yaw, and roll directions. When the head rotates rapidly, the extraocular muscles need to adjust the line of sight synchronously, resulting in the ciliary muscle bearing an additional co-accommodative load. By analyzing the correlation between the rate of change in head posture and the rising slope of the ciliary muscle electrical signal, G1 was determined as the critical value at which "head movements begin to significantly affect the ciliary muscle's accommodative efficiency"—that is, when G... H At G1, a decrease in electromyographic signal response speed exceeding 15% indicates a need for acceleration compensation to offset the delay. For example, if a student suddenly turns their head to look at the blackboard on the side of the classroom, and the rate of change of pitch angle is 40° / s, yaw angle is 50° / s, and roll angle is 20° / s, the combined motion gradient exceeds the normal values. This is considered a significant head movement, triggering an emergency event candidate condition.

[0087] The curvature of the three-dimensional trajectory quantifies the intensity of gaze shifts. When the gaze trajectory makes a sharp turn (such as quickly scanning from the corner of a desk to the edge of the blackboard), the ciliary muscle needs to complete asymmetrical adjustment in a short time. At this time, the curvature value exceeds the critical value K1, indicating that the visual system's demand for sharpness has increased sharply. Traditional compensation algorithms are prone to causing blurred vision due to response delays. The threshold is determined by fitting the curvature distribution of a large number of real gaze trajectories, and the value is selected to recognize 90% of intense gaze shift scenarios. For example, when students are reading textbook text and images, if their gaze quickly changes direction within a distance of 0.3m, the curvature of the resulting local trajectory exceeds K1, indicating a high-frequency, small-range adjustment demand, which needs to be triggered by the corresponding compensation mechanism.

[0088] By analyzing electromyographic signals under different regulatory scenarios, it was found that when Below a certain critical value, the ciliary muscle's ability to respond to rapid changes in distance significantly decreases, leading to a prolonged accommodative lag time; k th Defined as the minimum slope value at which the electromyographic signal response speed is sufficient to support routine adjustment needs. For example, by collecting ke data from over 200 adolescent users at different gaze switching speeds, the distribution characteristics of ke in emergency scenarios were fitted, and the 75th quantile was taken as k. th This ensures coverage of 75% of efficient adjustment and response scenarios.

[0089] Based on the rate of change of distance V D Head motion gradient G H The three-dimensional trajectory curvature κ and the slope ke of the ciliary electromyography signal are used to determine the formation of a dual verification mechanism of "behavioral characteristics + physiological indicators" in the model. D G H κ identifies high-load conditioning scenarios from the perspective of gaze behavior, while ke confirms the risk of decreased conditioning efficiency from the perspective of muscle response. Only when behavioral characteristics indicate high load ( , , And physiological indicators show that the muscle response is sufficient ( >k th Only when the condition is met (e.g., when the muscle response is good during a high-speed switch) is it considered an emergency regulatory event; otherwise, it is considered a stable regulatory event. This avoids overcompensation caused by misjudgment based solely on behavioral characteristics (e.g., not triggering unnecessary acceleration compensation when the muscle response is good during a high-speed switch). th The introduction of this technology fills the gap in traditional judgment models that "rely solely on external behavioral data and ignore individual muscle differences." For example, users with high myopia may experience muscle strain due to prolonged tension in the ciliary muscle. The values ​​are too low, even if the behavioral characteristics meet the standards. ≤k th It is still classified as a stable event and the compensation time constant is extended to avoid blurred vision caused by premature compensation due to insufficient muscle response.

[0090] Step S2100 solves the problem of inaccurate event recognition by fusing and jointly determining multimodal features; distance change rate The speed at which the ciliary muscle switches between near and far distances directly reflects the core driving force of ciliary muscle accommodation and is the most direct factor triggering the need for accommodation; head movement gradient It supplements the additional load from head movements, identifies the need for coordinated adjustment caused by neck movements, and avoids the missed detections caused by traditional solutions that only focus on distance changes; trajectory curvature κ By capturing the dynamic characteristics of gaze shifts, it identifies sudden accommodative demands in non-linear fixation scenarios (such as rapid saccadic gaze). While the distance change in these scenarios is small, the instantaneous response of the ciliary muscle is extremely demanding. Ke verifies the accommodative state at an intrinsic physiological level, ensuring that the judgment results accurately reflect individual muscle response capabilities. The combined judgment of these four factors forms a three-dimensional judgment model of "driving force - synergistic load - dynamic trajectory - physiological response," effectively improving the accuracy of accommodative event recognition compared to single-feature judgment.

[0091] Step S2200: If it is determined to be an emergency adjustment event, then the advanced compensation acceleration mechanism is triggered.

[0092] Specifically, when step S2130 determines an emergency adjustment event, the advanced compensation acceleration mechanism is triggered, which includes the following sub-steps: First, based on the real-time distance change rate V in the dual-modal sensing data... D Head motion gradient G H The three-dimensional trajectory curvature κ and the slope of the ciliary muscle electrical signal change ke are used to identify the characteristics of the current gaze switching, such as rapid distance changes, significant head movements, sharp trajectory turns, and violent fluctuations in ciliary muscle electrical signals. In such scenarios, the ciliary muscle accommodation load increases sharply, and traditional fixed-delay compensation strategies are prone to causing severe lag in refractive compensation.

[0093] To address this issue, the advance compensation acceleration mechanism first shortens the response period of lens refractive compensation to match the dynamic characteristics of emergency events, for example, by increasing the data processing frequency to twice the original sampling frequency (e.g., from 100Hz to 200Hz) to reduce signal processing delay. Secondly, a feedforward prediction model based on the slope ke of the ciliary electromyography signal is introduced: by analyzing the correlation between ke and accommodative hysteresis in historical data, a real-time mapping function is established when... Exceeding the preset threshold k th Based on the current value of ke, the model predicts the trend of accommodative demand changes within the next 100ms and adjusts the lens refractive power in advance. For example, if the current ciliary muscle electrical signal rises at a slope of 50μV / s (significantly higher than the average slope of 10μV / s in a stable scene), the predictive model determines that the ciliary muscle is rapidly contracting to cope with near fixation, and thus applies a refractive compensation increment of +0.5D in advance, responding 200ms earlier than traditional delayed compensation.

[0094] In addition, regarding the head motion gradient G H In cases of high elevation, the compensation mechanism simultaneously introduces a head motion compensation coefficient: this coefficient is based on G... H The real-time value is dynamically adjusted when G H When the G1 threshold is exceeded, an adjustment correction related to the head movement direction and speed is superimposed on the original compensation algorithm. For example, when the head tilts upward rapidly (accompanied by a switch in distant gaze at the blackboard), the adjustment is based on the yaw angle change rate. and pitch angle change rate The vector synthesis results are supplemented with an additional 0.3D of hyperopia compensation to counteract the synergistic load of extraocular muscle movement on ciliary muscle accommodation.

[0095] The proactive compensation acceleration mechanism, through deep fusion of dual-modal data, accurately captures the sudden demands of the ciliary muscle during emergency accommodation events, solving the compensation lag problem caused by fixed delays in traditional solutions. For example, when a student quickly looks up from their desk (0.3m) to the blackboard (5m) with a head pitch rate of 30° / s (exceeding the threshold of 20° / s), the proactive compensation acceleration mechanism can complete the refractive power adjustment in advance, in a shorter time than traditional solutions. This effectively reduces instantaneous refractive errors caused by accommodation lag, avoids the accumulation of ineffective accommodation movements, and alleviates excessive fatigue of the ciliary muscle in high-frequency switching scenarios.

[0096] In step S2300, if the event is determined to be a stable adjustment event, a conventional compensation mechanism is adopted.

[0097] Specifically, the conventional compensation mechanism is based on the average accommodation delay in historical statistics, constructing a steady-state compensation model: It smooths the gaze distance, head posture data, and ciliary electromyography signals using a sliding time window (e.g., a 1-second window), calculating the mean of the data within the window as the input parameter for the current accommodation state. For example, when the gaze distance slowly switches between 0.5m (first region) and 3m (second region), and the distance change rate V... D When the velocity is 0.8 m / s (below the threshold of 1.5 m / s), the system uses the average gaze distance (2.0 m) of 20 consecutive sampling points within the window (sampling interval 50 ms) as the basis for calculating the target refractive power, so as to avoid the interference of a single noise point on the compensation strategy.

[0098] Secondly, a baseline drift correction module for ciliary muscle electromyography (CEMG) signals is introduced into the conventional mechanism: Since the slope of CEMG signals is small in stable scenarios, they are easily affected by noise such as fluctuations in skin impedance. By calculating the difference between the current signal and the baseline signal from the previous n² seconds, the effective component reflecting the true accommodative state is extracted, ensuring that the refractive compensation amount matches the actual physiological needs. For example, when the baseline signal is 20μV and the current signal is 25μV, the ciliary muscle is determined to be in a state of mild contraction, corresponding to 0.75D myopia compensation, rather than directly using the original signal, which could lead to misjudgment (such as misjudging as 1.5D compensation corresponding to 30μV under noise interference).

[0099] The conventional compensation mechanism addresses the physiological characteristics of stable accommodation scenarios by reducing system computational complexity while maintaining compensation accuracy through data smoothing and baseline correction. For example, when a student slowly moves their gaze in the first region (e.g., slightly turning their head while reading a textbook), the conventional compensation mechanism can control data processing latency to within 150ms. Simultaneously, baseline correction reduces noise interference from ciliary muscle electroencephalogram (CME) signals by 60%, avoiding frequent refractive power adjustments caused by over-response to low-frequency signal fluctuations (e.g., reducing ineffective adjustments from 5 per minute in traditional solutions to 1), thus improving visual stability and wearing comfort in dynamic visual scenarios. Furthermore, the conventional compensation mechanism complements the advanced compensation mechanism, covering the entire range of fixation switching scenarios from low-frequency stability to high-frequency urgency. This ensures that the lens refractive compensation strategy is synchronized with the ciliary muscle accommodation rhythm under different loads, reducing the overall amplitude of abnormal refractive power fluctuations and achieving a balance between physiological adaptability and compensation accuracy.

[0100] The outcome of an emergency event directly drives the proactive compensation acceleration mechanism, which pre-adjusts the lens refractive power to offset physiological delays and resolves the problem of amplified instantaneous errors during high-speed switching. The outcome of a stable event triggers the conventional compensation mechanism, ensuring visual comfort through gradual adjustment and avoiding user discomfort caused by frequent and rapid adjustments. This hierarchical decision-making mechanism maximizes the value of the fusion of dual-modal perception data (physiological signals and spatial motion data), avoiding the misjudgment risk of a single modality (such as relying solely on distance changes) and achieving "on-demand compensation" through dynamic parameter adjustment. This fundamentally solves the core problem of asynchronous compensation strategies and physiological responses in existing designs, providing crucial technical support for the accurate adaptation of myopia control lenses in real-world dynamic visual scenarios.

[0101] Example 2

[0102] This embodiment, based on Embodiment 1, provides a refractive compensation method based on dual-modal sensing, including:

[0103] Step S1000: Establish the head-mounted device body coordinate system O-XYZ, and define the first region and the second region in the coordinate system O-XYZ; the user performs back-and-forth gaze switching between the first region and the second region at different switching speeds, collects bimodal perception data during the user's gaze switching process, and obtains the individual initial compensation coefficient matrix A.

[0104] Further, step S1000 includes:

[0105] Step S1100: Establish the head-mounted device body coordinate system O-XYZ, and define the first region and the second region in the coordinate system O-XYZ;

[0106] The method for establishing the head-mounted device body coordinate system O-XYZ and defining the first and second regions in the coordinate system O-XYZ is the same as in Embodiment 1.

[0107] Step S1200: Based on the head-mounted device body coordinate system O-XYZ, perform personalized parameter initialization to obtain the individual initial compensation coefficient matrix A;

[0108] Further, step S1200 includes:

[0109] Step S1210: Based on the head-mounted device body coordinate system O-XYZ, the user is required to perform N1 round-trip gaze switching between the first and second regions at different switching speeds, with each speed repeated m times.

[0110] Step S1220: Simultaneously acquire dual-modal perception data and adjust delay parameters during the user's gaze switching process. ;

[0111] The dual-modality refers to the physiological signal modality and the spatial motion modality. The dual-modal sensing data includes physiological signal data and spatial motion data. The physiological signal data refers to ciliary electromyography (CEMG) signals. The spatial motion data includes the user's head posture data, gaze distance, and three-dimensional gaze point coordinates. A global spatial curve is generated from the three-dimensional gaze point coordinates.

[0112] The method for collecting the user's head posture data is as follows: a nine-axis inertial sensor is integrated into the bridge of the nose of the glasses frame to collect head posture data; the head posture data includes the head pitch angle. Yaw angle Roll angle ,as well as , and rate of change ;

[0113] The method for acquiring the ciliary muscle electromyography (CEMG) signal is as follows: an n1-channel dry electrode bioelectric sensor is arranged on the inner side of the temple of the spectacle to acquire the CEMG signal.

[0114] Specifically, users perform round-trip gaze switching between the first and second zones at different switching speeds. For example, in a typical scenario where students frequently switch between a near desk (approximately 0.3-0.5m) and a far blackboard (approximately 2-10m), multiple switching speeds are set, such as low speed (gaze distance change rate ≤ 0.5m / s), medium speed (0.5m / s < change rate < 1.5m / s), and high speed (change rate ≥ 1.5m / s), covering different behavioral patterns in daily learning, from slow head tilting to rapid scanning. Each speed level is repeated m times (e.g., m=10) to reduce the interference of individual accidental movements on data collection and ensure statistical significance of the samples. In the head-mounted device's coordinate system, users complete a round-trip gaze from the first zone to the second zone along a preset path, i.e., shifting from a near target (such as text in a textbook) to a far target (such as writing on a blackboard) and back to the near target, forming a complete gaze cycle. Head posture and gaze point coordinates are monitored in real time using a nine-axis inertial sensor and a ToF sensor. By conducting repeated tests at multiple speeds and with multiple repetitions, a dataset containing individual dynamic accommodation characteristics was constructed, providing a foundation for establishing a mapping relationship between "gaze speed - physiological response - compensation needs." Without switching tests at different speeds, the compensation strategy will be unable to distinguish between rapid salivation and slow accommodation scenarios, leading to the accumulation of compensation delays during high-frequency switching and exacerbating the ciliary muscle's accommodation load.

[0115] An n1-channel dry electrode bioelectric sensor (e.g., n1=4, symmetrically distributed on both temples) is placed on the inner side of the temples to non-invasively acquire ciliary electromyography (EMG) signals using surface electrode technology. The dry electrodes do not require conductive gel, making them suitable for children and reducing skin irritation. The raw signal is bandpass filtered (10-500Hz) to remove power frequency interference (e.g., 50Hz mains noise) and DC drift, and then baseline correction is used to eliminate long-term trends, ensuring the accuracy of subsequent feature extraction.

[0116] The head pitch angle θ (rotation around the X-axis, positive when head is tilted down), yaw angle ϕ (rotation around the Y-axis, positive when turning left), roll angle ψ (rotation around the Z-axis, positive when tilting right), and the rate of change of each angle are collected by a nine-axis inertial sensor integrated in the bridge of the nose (including a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer). The ToF sensor reflects the direction and speed of head movement. The distance to the gaze target is measured in real time using a ToF sensor, and the coordinates of the gaze point in three dimensions are determined by combining this with the coordinate system of the head-mounted device. Connecting consecutive gaze points generates a global spatial curve for subsequent trajectory curvature calculation.

[0117] Single adjustment delay time τ i' , i'=1,2,...,N1: During each fixation switch, the time difference between the moment the fixation distance begins to change (i.e., the instant the Z-axis coordinate exceeds the current region threshold, such as the instant the change occurs from the first region Z≤0.6m to the second region Z≥2m) and the moment the ciliary muscle electrical signal reaches its peak. This time difference reflects the physiological delay in the ciliary muscle from receiving the visual distance change signal to completing contraction regulation.

[0118] Adjusting delay parameters Determination of τ for N1 round-trip gaze switching i' (i'=1,2,…,N1) Take the arithmetic mean to obtain , The average time lag of the ciliary muscle accommodation response relative to lens compensation under normal fixation switching is used as the initial benchmark for time parameters in subsequent compensation strategies.

[0119] By simultaneously acquiring dual-modal data, a correlation is established between "spatial motion trajectory → physiological regulatory signals → delay parameters," avoiding the disconnect between compensation strategies and physiological responses caused by traditional methods that rely solely on single-modal data (such as fixation distance). Adjusting the delay parameters... As a basic physiological characteristic of an individual, it is used to correct the delay time of the advanced compensation acceleration mechanism in step S2300, ensuring that the timing of compensation is synchronized with the ciliary muscle regulation rhythm.

[0120] Without the bimodal data acquisition in this step, the subsequent individual initial compensation coefficient matrix A will lack coupling information between physiological signals and spatial motion. This will cause the accommodative lag prediction model to fail to accurately map the relationship between "gaze trajectory change - ciliary electromyography characteristics - compensation demand," ultimately resulting in a disconnect between the compensation strategy and the actual physiological response, and failing to address the problem of ineffective accommodative load in high-frequency switching scenarios. By synchronously acquiring bimodal data in this step, individual physiological characteristics (ciliary electromyography signals) and spatial motion behaviors (head posture, gaze trajectory) are quantitatively correlated, providing an indispensable data source for the precision of subsequent dynamic compensation strategies. This is a key technical step in achieving "physiologically adaptive refractive compensation."

[0121] Step S1230: Based on the bimodal perception data during the user's gaze switching process, obtain the individual initial compensation coefficient matrix A.

[0122] Furthermore, such as Figure 5 As shown, step S1230 includes:

[0123] Step S1231: Spatiotemporal alignment of the user's head posture data, ciliary electromyography signals, and gaze distance;

[0124] Transform the user's head posture data, ciliary electromyography signals, and gaze distance into the O-XYZ coordinate system of the head-mounted device;

[0125] Step S1232: Feature extraction is performed on the spatiotemporally aligned ciliary electromyography (EMG) signal to obtain EMG response features reflecting the user's ciliary muscle contraction activity; the EMG response features include EMG energy E EMG ;

[0126] Step S1233: Based on the spatiotemporally aligned head posture data, gaze distance, and electromyographic response characteristics, the refractive compensation coefficient is calculated using a multivariate regression model to form an individual initial compensation coefficient matrix A.

[0127] Specifically, step S1230 aims to construct an individual initial compensation coefficient matrix A by processing and modeling bimodal sensing data, establish a quantitative mapping relationship between "physiological signal-spatial location-compensation demand", and solve the technical problem of the disconnect between compensation strategy and real physiological response in existing designs.

[0128] In step S1231, the midpoint of both pupils is taken as the origin, the X-axis extends horizontally to the right along the line connecting the pupils, the Y-axis points vertically to the top of the head, and the Z-axis extends along the line of sight directly in front (perpendicular to the lens plane), ensuring that all spatial motion data (such as the coordinates of the gaze point) are strictly aligned with the physical position of the device. The output data from the nine-axis inertial sensor (collecting head posture data), dry electrode sensor (collecting electromyographic signals), and ToF sensor (collecting gaze distance) are timestamped. Linear interpolation is used to unify the asynchronous data from different sensors to a common sampling frequency, eliminating timing deviations caused by differences in sampling rates. Without spatiotemporal alignment, the spatial reference system and time base of different modal data will be inconsistent, leading to errors in subsequent feature association. For example, if the change in head pitch angle θ and the spatial position of the gaze distance are not correlated in the same coordinate system, it may misjudge the adjustment load when looking down at a desk, causing the compensation strategy to deviate from actual needs.

[0129] In step S1232, the spatiotemporally aligned ciliary electromyography (CEMG) signal undergoes preprocessing and feature calculation, including bandpass filtering (10-500Hz), using a Butterworth filter to remove power frequency interference (such as 50Hz mains noise) and DC drift, retaining the effective electrical signal generated by ciliary muscle contraction (typical frequency range 10-500Hz); baseline correction: polynomial fitting removes signal baseline drift, using the mean value of the resting state 2 seconds before signal acquisition as the baseline reference to ensure the accuracy of subsequent energy calculations. The root mean square (RMS) algorithm is used to calculate the electromyographic energy value of the preprocessed signal within a sliding time window (e.g., 50ms). Ciliary electromyography signals are a direct representation of the regulatory physiological state, but the original signal is susceptible to noise interference. Through filtering, baseline correction, and RMS calculation, the physiological signal is transformed into a quantifiable regulatory load index, providing a physiological driving basis for subsequent compensation strategies. For example, when a user quickly switches from the second region to the first region, the EEMG increases significantly, indicating that the ciliary muscle needs to contract rapidly to adapt to near vision. At this time, the refractive compensation force needs to be increased to counteract the accommodative lag.

[0130] In step S1233, the preprocessed bimodal data is grouped according to multidimensional features, and a linear regression model is fitted using the least squares method to generate an individual-specific refractive compensation coefficient matrix A. The specific steps are as follows:

[0131] The user executed in step S1210 The data on each round-trip gaze switching is grouped according to the following dimensions:

[0132] Head posture grouping: (0°-180°) and Discretization interval (0°-360°) (e.g.) Every Every Divide an interval into subsets;

[0133] Grouping of fixation distances: Discretized at 0.1m intervals according to the first and second regions, including typical fixation distances from the first region at near distance to the second region at far distance;

[0134] Electromyographic energy stratification: E of all samples EMG The ciliary muscle is divided into three levels—low, medium, and high—based on the 25th and 75th percentiles, reflecting different levels of ciliary muscle accommodative load.

[0135] For the data within each group, a linear regression model was fitted using the least squares method, with the refractive error compensation value as the dependent variable and E0 as the denominator. EMG Using the least squares method as the independent variable, a linear equation was fitted to obtain the refractive compensation coefficient. , Indicates a specific head posture gaze distance and ciliary muscle load Below, the lens refractive compensation value corresponding to a unit change in electromyographic energy (i.e., the refractive power compensation corresponding to each microvolt of electromyographic signal). i is the head pitch angle. θ Discretized interval index, j is the head yaw angle. ϕ The discretized interval index, k is the discretized point index of the gaze distance, and l is the electromyographic energy E. EMG Hierarchical index.

[0136] Regression coefficients for all groups according to Store and form the individual initial compensation coefficient matrix. .in, pitch angle The number of discretization intervals (e.g., 36, corresponding to) , interval); Yaw angle The number of discretization intervals (e.g., 72, corresponding to) interval); gaze distance The number of discretization points (e.g., 150, corresponding to) interval); For electromyographic energy The number of strata (e.g., 3, low / medium / high).

[0137] matrix By using regression analysis, the head pose data... gaze distance electromyography energy With refractive compensation coefficient This direct correlation forms a mapping relationship of "physiological signal - spatial location - compensation need". For example, when a user gazes at a distance of 0.3m (close distance) with θ=10∘ (looking down at a desk) and ϕ=0∘ (looking straight ahead), and E... EMG When at a higher level, the corresponding a in matrix A ijk,l This will instruct a greater refractive compensation level to counteract ciliary muscle accommodative lag. Because step S1210 requires the user to perform round-trip fixation switching at different speeds, the matrix... A This technology can capture the differences in individual regulatory responses during fast / slow switching, providing a personalized training data foundation for subsequent regulatory lag prediction models. Individual initial compensation coefficient matrix. A It not only includes the coupling relationship of dual-modal sensing data, but also explicitly expresses the quantitative relationship between ciliary electromyography signal intensity and refractive compensation through regression modeling, providing core benchmark data for subsequent dynamic compensation strategies and solving the technical problem of "compensation strategy not being synchronized with real physiological response".

[0138] Step S1230 unifies head posture, gaze position, and electromyography (EMG) signals into a unified spatiotemporal reference through spatiotemporal alignment, resolving the disconnect between physiological signals and spatial motion data in traditional solutions. The modal coupling mechanism enables the compensation strategy to possess dual regulatory capabilities of "physiological signal-driven + spatial position awareness," significantly improving adaptability in dynamic scenarios. EMG energy feature extraction and multivariate regression modeling transform the ciliary muscle regulatory load into calculable compensation parameters. Traditional solutions rely on fixed delay parameters, failing to reflect the impact of EMG intensity changes on compensation requirements; however, this step quantifies EMG using the RMS algorithm and uses regression coefficients... A linear mapping is established to dynamically adjust the compensation intensity according to the electromyographic load. Matrix A serves as the mapping hub between "physiological signal-spatial location-compensation demand," connecting initial data acquisition with subsequent regulatory lag prediction and compensation mechanisms. Through multi-velocity data acquisition and hierarchical modeling, Matrix A can capture the specific regulatory demands during rapid switching. When a user moves rapidly between the first and second regions, their corresponding grouping will trigger a greater... This prompts the lens to adjust its refractive power in advance, offsetting the physiological delay during rapid switching. This refined grouping strategy upgrades the compensation system from "fixed parameter response" to "dynamic feature matching," effectively reducing abnormal fluctuations in refractive power and alleviating high-frequency accommodative fatigue of the ciliary muscle.

[0139] Step S2000: Based on the bimodal perception data during the user's gaze switching process, determine whether it is an emergency adjustment event or a stable adjustment event; if it is determined to be an emergency adjustment event, trigger the advance compensation acceleration mechanism; if it is determined to be a stable adjustment event, adopt the conventional compensation mechanism.

[0140] Further, step S2000 includes:

[0141] Step S2100: Based on the bimodal perception data during the user's gaze switching process, determine whether it is an emergency adjustment event or a stable adjustment event;

[0142] Specifically, it is defined as the minimum interval ΔD between the first region and the second region. th When the gaze distance switches from the first region to the second region or vice versa, the actual change range is... It needs to exceed ΔD th Only then can the event determination be triggered. Among them, The current viewing distance. The gaze distance at the previous time step t is the distance at the current time t. This is determined by statistically analyzing the average rate of change of distance at different switching speeds during step S1210 (i.e., ...). The average rate under high-frequency switching scenarios (such as more than 2 switching per second) is taken as the distance change rate threshold. (For example ). Calculate the rate of change of electromyographic energy. ,in, This represents the current electromyographic energy. The electromyographic energy from the previous moment; The threshold of the rate of change of electromyographic energy reflects the instantaneous change in the ciliary muscle's accommodative load. Set as the average level of electromyographic energy mutation during rapid switching in an individual (e.g. ).

[0143] A dual-condition joint determination mechanism is adopted:

[0144] Emergency Adjustment Event Determination: An emergency adjustment event is determined when both of the following conditions are met simultaneously:

[0145] (1) Variation in gaze distance and the rate of change (This indicates that the gaze point switches rapidly between the first and second regions, such as a student quickly looking up at the blackboard or looking down to take notes as the teacher explains something.)

[0146] (2) Rate of change of electromyographic energy (This indicates that the ciliary muscle experiences significant load changes due to rapid regulation, such as a sharp increase in electromyographic signals within a short period of time, reflecting a sudden increase in regulatory demand.)

[0147] Determination of stationary adjustment events: If neither of the above two conditions is met simultaneously, it is determined to be a stationary adjustment event, including the following scenarios:

[0148] (1) The range of change in fixation distance is less than (e.g., slightly shifting the line of sight within the first area) Axis coordinates in [ (within the range of variation)

[0149] (2) The rate of change of distance is lower than (For example, if you slowly look up at the blackboard, the ciliary muscle has enough time to adjust.)

[0150] (3) The rate of change of electromyographic energy is lower than (This indicates that the ciliary muscle's regulatory load has not undergone a significant change and is in a normal regulatory state.)

[0151] This determination method solves the key problem of "how to identify sudden changes in ciliary muscle accommodation load in real time" through joint analysis of dual-modal data. Emergency accommodation events correspond to scenarios with a high risk of ciliary muscle accommodation lag (such as rapid switching between near and far vision, where physiological response delays can easily lead to the accumulation of instantaneous refractive errors). By identifying such events in advance, the advanced compensation acceleration mechanism in step S2300 can be triggered, shortening the compensation delay time. This allows the lens refractive power to match the target distance requirement in advance, avoiding the accumulation of ineffective accommodative load. Without this judgment step, the system cannot distinguish the urgency of the accommodative event, potentially leading to the continued use of the conventional compensation cycle during rapid switching, exacerbating the asynchrony between refractive compensation and physiological response, and failing to effectively alleviate ciliary muscle fatigue. By combining spatial motion data (distance change amplitude / rate) with physiological signal data (electromyographic energy change rate), a mapping relationship of "scene characteristics - physiological response" is established, enabling the individual initial compensation coefficient matrix collected in step S1000 to... A (Including compensation coefficients under different head postures, gaze distances, and electromyographic loads) can be dynamically correlated with real-time adjustment events. For example, when an emergency event is determined, the system can prioritize calling the matrix. A The corresponding high electromyographic energy stratification ( l =3) and a compensation coefficient for rapid distance changes, enhancing the targeting of the compensation strategy. Personalized thresholds (based on individual historical data) , , This allows it to adapt to differences in the regulatory abilities of different users (such as the different ciliary muscle reaction speeds between children and adults), avoiding the shortcomings of traditional fixed-parameter algorithms in terms of individual adaptability. For example, if a user exhibits an average electromyographic energy change rate of 40% during rapid switching in step S1210, then their... It can be set to 35% to ensure that emergency compensation is triggered only when the load significantly exceeds its normal regulation load, reducing the probability of misjudgment.

[0152] The coordinate system defined in step S1000 O - XYZ This provides a spatial positioning reference for calculating the gaze distance. The dual-modal data collected in step S1210 at different switching speeds serves as the threshold setting (e.g., , This provides a basis for understanding individual differences. The judgment result (emergency / stable event) serves as a prerequisite for the accommodation lag prediction model (see step S2200), enabling the model to invoke different compensation strategies (advanced compensation or conventional compensation) based on the event type. This ensures the dynamic matching of "physiological signal characteristics - spatial motion trajectory - compensation strategy," ultimately achieving synchronization between lens refractive compensation and ciliary muscle accommodation rhythm, thus resolving the core defect in the original technology where "compensation strategies are not synchronized with actual physiological responses." Through the above judgment method, high-risk accommodation scenarios can be identified in real time, providing a decision-making basis for the differentiated execution of subsequent compensation mechanisms. This forms a complete closed loop from data collection and feature analysis to strategy response, significantly improving the fitting accuracy and physiological compatibility of myopia control lenses in dynamic visual scenarios.

[0153] Step S2200: Based on the individual initial compensation coefficient matrix A, construct a regulation lag prediction model to obtain the predicted regulation lag ΔF. pred ;

[0154] Further, step S2200 includes:

[0155] Step S2210, based on the individual initial compensation coefficient matrix A and the three-dimensional trajectory curvature Head motion gradient G H Distance change rate V D A regulatory hysteresis prediction model was constructed based on the slope ke of the change in ciliary muscle electromyography signal.

[0156] Step S2220: Based on the constructed adjustment lag prediction model, obtain the predicted adjustment lag amount ΔF. pred .

[0157] Specifically, step S2200 aims to construct a ciliary muscle accommodative lag prediction model. By fusing multi-dimensional sensory data, it dynamically predicts the amount of ciliary muscle accommodative lag, thereby providing a real-time correction basis for lens compensation strategies. This step solves the core problem of traditional compensation algorithms relying on fixed delay parameters, which leads to error amplification, and achieves synchronization between refractive compensation and physiological accommodation rhythm. Individual initial compensation coefficient matrix A, three-dimensional trajectory curvature... Head motion gradient G H Distance change rate V D The method for obtaining the slope ke of the ciliary electromyography signal change is the same as in Example 1. The adjustment lag prediction model adopts a temporal convolutional neural network (TCN) architecture based on the attention mechanism. The model input layer contains four parallel branches: (1) the vector of the individual initial compensation coefficient matrix A after flattening, which represents the user's personalized adjustment characteristics; (2) the time series of the three-dimensional trajectory curvature κ, which reflects the abrupt change characteristics of the gaze trajectory; (3) the head motion gradient G. H The mean of the sliding window represents the intensity of head movement; (4) the rate of change of distance VD The differential sequence captures the dynamic characteristics of gaze switching. After temporal alignment, each branch's data is extracted using a one-dimensional convolutional layer. An attention mechanism module dynamically weights and fuses the features from each branch, highlighting the dominant influencing factor at the current moment. A fully connected layer maps the fused features to a predicted adjustment lag ΔF. pred During training, the actual adjustment delay time τ collected in step S1200 is used. i' As a supervisory signal, time-series cross-validation is used to divide the training and test sets, and Huber loss is used to balance the impact of outliers. During the model deployment phase, the slices of matrix A, κ, and G at the current time step are input in real time. H V D And historical time series window data, output ΔF within the future time window Δt''. pred Predicted value. Predicted adjustment lag ΔF pred This represents the instantaneous deviation between the actual accommodative force of the ciliary muscle and the target refractive demand. This value is obtained by inverse quantization of the normalized value output by the TCN model. During real-time prediction, the model updates ΔF every 50ms. pred This ensures the timeliness of the compensation strategy.

[0158] Traditional fixed delay parameter τ A The model cannot reflect the nonlinear regulatory characteristics of individuals during rapid gaze switching. By introducing multidimensional dynamic feature modeling, the accuracy of hysteresis prediction is significantly improved. The spatiotemporal correlation between head movement and gaze trajectory affects the regulatory hysteresis pattern, and the temporal modeling capability of TCN effectively captures such complex correlations. Furthermore, the individual's initial compensation coefficient matrix A provides personalized prior knowledge for the model, avoiding the slow convergence problem caused by learning from scratch. The multi-velocity switching data collected in step S1210 provides rich training samples for the TCN model, enabling the model to learn ΔF at different speeds. pred With V D The nonlinear mapping relationship; the matrix A generated in step S1233 is used as the model input, and the personalized features calibrated offline are embedded into the real-time prediction process to improve the model's convergence speed and generalization ability. This collaboration ensures the accuracy of the baseline while enabling a rapid response to sudden adjustment events.

[0159] If this step is missing, it will lead to three technical obstacles: (1) The advance compensation acceleration mechanism of step S2300 lacks quantitative correction basis and cannot dynamically adjust τ. realThis leads to inaccurate compensation timing in emergency situations; (2) the individual initial compensation coefficient matrix A generated in step S1233 only contains static correlations and cannot adapt to real-time changes in regulatory needs, resulting in a disconnect between compensation strategies and dynamic physiological responses; (3) the online update mechanism in step S2500 loses its benchmark model and cannot achieve continuous optimization of compensation parameters. This step establishes an interpretable prediction model to integrate personalized static parameters (matrix A) with dynamic sensing data (κ, G). H V D The organic combination of these elements enables the lens compensation system to have adaptive evolution capabilities.

[0160] When the user performs frequent gaze switching, the TCN model relies on the real-time head motion gradient G. H To determine the degree of head rotation, when G... H When the value exceeds G1, the model automatically increases the weighting coefficient of the 3D trajectory curvature κ to promptly capture sudden changes in adjustment needs caused by abrupt changes in gaze. For example, when a student quickly turns their head to look at the side blackboard, the increased κ value triggers the model to output a higher ΔF. pred The driving lens increases the positive refractive compensation amount in advance to counteract the accommodative lag caused by sudden head turns. Simultaneously, the personalized compensation coefficients stored in matrix A ensure that different users experience the same G-force. H The differential compensation amount that adapts to its physiological characteristics is obtained when the κ value is obtained.

[0161] Step S2300: If an emergency adjustment event is determined, an advance compensation acceleration mechanism is triggered; the advance compensation acceleration mechanism dynamically shortens the adjustment delay parameter. The shortened compensation delay time is obtained. ,based on Calculate the target refractive power in advance ,according to Adjust the lenses.

[0162] The shortened compensation delay time The method is as follows: Obtain the electromyographic energy at the current moment. Based on the current electromyographic energy Shorten the adjustment delay parameter To obtain compensation for the delay time .

[0163] The calculated target refractive power The method is to obtain the current gaze distance. The distance of gaze from the previous moment ,according to , and the predicted adjustment lag ΔF pred Determine the target refractive power .

[0164] According to The method for adjusting the lenses is as follows: Generate motor angle control commands to fully rotate the lens and achieve dynamic adjustment of refractive power.

[0165] Specifically, ciliary electromyography (EMG) signals are acquired in real time using dry electrode bioelectric sensors placed on the inner side of the temple. After bandpass filtering (10-500Hz) and baseline correction, the EMG energy within the current time window is calculated using the root mean square (RMS) algorithm. This energy value reflects the instantaneous contraction intensity of the ciliary muscle during an emergency regulatory event; higher energy indicates a greater load on the ciliary muscle due to regulatory lag. Based on the current electromyographic energy... With preset energy threshold The ratio is used to adjust the delay parameter according to the following rules:

[0166] like ,but ,in Shortening factor This represents the maximum historical electromyographic energy for an individual.

[0167] like ,but .

[0168] It is a dimensionless proportionality coefficient used to account for the excess portion of electromyographic energy. This is mapped to the amount of time reduction in delay. Its value range is... To ensure the shortened delay time Always less than the original delay parameter However, it will not be shortened excessively to zero to avoid compensating for refractive power oscillations caused by premature triggering.

[0169] lower limit This ensures that when electromyographic energy exceeds the threshold, the compensation delay time will inevitably be shortened, demonstrating the necessity of "preemptive compensation".

[0170] upper limit Preventing electromyographic energy from reaching its maximum value hour, When the lens compensation is reduced to zero, it causes the lens compensation to completely deviate from the physiological regulatory rhythm, leading to overcompensation (such as dizziness caused by sudden changes in refractive power).

[0171] This step advances the compensation timing by linearly linking high electromyographic energy to the reduction in delay time. For example, when a user quickly switches from the second region to the first region, if the detection... If it is significantly higher than the energy threshold, then it shortens. This allows for early triggering of lens refractive compensation, thereby offsetting the accommodative lag caused by the ciliary muscle's rapid switching.

[0172] according to , and the predicted adjustment lag ΔF pred Determine the target refractive power The formula is ;in, Used to determine the direction of gaze switching (negative for far-to-near, positive for near-to-far). By introducing... For ideal refractive power The correction is made to counteract the physiological delay of the ciliary muscle, so that the refractive power of the lens matches the target distance requirement in advance.

[0173] The frame integrates a dual-lens flip-up module containing two independently driven refractive lenses (L1 and L2), each corresponding to a different refractive power (e.g., L1 for +2.0D and L2 for -2.0D). Each lens is driven by a miniature servo motor, achieving angle adjustment within a 180° range via a rack and pinion mechanism. Target refractive power. The process of converting commands into motor angle control instructions is as follows:

[0174] Establish the "angle-refractive power" mapping relationship: According to the optical principle of double lens superposition, the target refractive power is determined by the effective angle of the two lenses in the optical path. For example, L1 is fully connected when it is 0°, L2 is fully connected when it is 90°, and the intermediate angle is continuously adjusted by combining trigonometric functions.

[0175] Drive strategy: When an emergency is detected, the motor quickly rotates to the target angle with maximum acceleration, and at the same time activates the dual-lens synchronous calibration mode. The built-in Hall sensor provides real-time feedback on the lens position to ensure independent adjustment for the left and right eyes (suitable for users with anisometropia).

[0176] The aforementioned mechanical control scheme achieves distortion-free refractive adjustment through physical optical switching. For example, for users with differences in diopter between their eyes, the left and right lenses can be independently adjusted to different angles, providing personalized compensation. A single lens structure would not be sufficient to meet the precise adjustment needs of anisometropia (different refractive errors).

[0177] Step S2300 solves the problem that traditional fixed delay parameters cannot adapt to rapid distance switching in emergency adjustment events. This is achieved by dynamically shortening the delay time ( ) and combined with the predicted lag ( This step synchronizes lens compensation with the actual response rhythm of the ciliary muscle. For example, in classroom scenarios where users frequently switch their gaze targets, without this step, lens compensation would lag behind physiological accommodation needs, leading to the accumulation of ineffective accommodative movements, exacerbating refractive fluctuations and visual fatigue. This step, however, adjusts the compensation timing in real time through electromyographic energy feedback, ensuring the lens reaches the target refractive power before the ciliary muscle completes accommodation, reducing ineffective load and significantly improving dynamic visual acuity. The independent driving of dual lenses solves the problem of anisometropia fitting, broadening the applicable user group (such as strabismus patients).

[0178] Step S2400: If the event is determined to be a stable accommodation event, a conventional compensation mechanism is adopted; based on the current gaze distance. Periodic calculation of refractive compensation ,according to Adjust the lenses.

[0179] Specifically, the conventional compensation mechanism triggers lens refractive power adjustment at fixed time intervals. The length of these time intervals is determined by the compensation response time corresponding to the discretized intervals of pitch angle θ and yaw angle φ in the individual's initial compensation coefficient matrix A. Specifically, under stable accommodation events, the user's head posture and gaze trajectory change relatively smoothly, and the ciliary muscle's accommodation demand does not exceed the normal physiological response range. In this case, the lens compensation strategy adjusts the gaze distance D at a predetermined time period (e.g., every 200 milliseconds). t Sampling is performed, and the corresponding refractive compensation coefficient a is retrieved from the individual initial compensation coefficient matrix A by combining the discretized indices of θ and φ in the current head pose data. ijk,l This coefficient is then weighted and calculated with the current electromyographic energy to obtain F. normal During weighted calculation, different weighting factors are assigned based on the stratification of electromyographic energy (low, medium, high) to dynamically adapt to the intensity differences in ciliary muscle regulatory load. If F normalIf the actual refractive error difference corresponding to the current lens angle exceeds a preset refractive threshold (e.g., 0.1D), a smooth switching mode is triggered: the motor rotates at a constant speed to the target angle, avoiding mechanical wear caused by high-frequency micro-adjustments. For example, when switching from 0.5D compensation to 0.7D compensation, the motor rotates at a constant speed, reducing the impact load on the rack and pinion mechanism; in non-switching states, the motor enters a "zero-torque standby mode," keeping only the angle encoder operational, reducing system energy consumption. The angle encoder is a high-precision position sensor integrated inside the micro servo motor, used to monitor the rotation angle of the lens module in real time, providing closed-loop position feedback for the lens adjustment system. This strategy solves the problem of wasted hardware resources caused by over-reliance on dynamic prediction models in stable scenarios. For example, in slow reading scenarios, the system samples at a fixed period, avoiding high-frequency calculations of real-time prediction models and extending device battery life. Without periodic control, even in low-load scenarios, the system may still frequently adjust the lens, leading to shortened motor life and increased power consumption.

[0180] Step S2400 transforms complex real-time prediction into parameter retrieval and weighted calculation through a fast lookup mechanism in matrix A, significantly reducing the processor's computational load. The smooth switching mode avoids frequent motor starts and stops and rapid acceleration, and combined with the zero-torque standby mode, it can extend the lifespan of mechanical components. In conjunction with the online update mechanism in step S2500, periodically compensated data provides samples for the incremental learning of matrix A, enabling it to adapt to long-term changes in individual regulatory habits (such as improved regulatory capacity due to ciliary muscle development in children).

[0181] Step S2500: After each refractive compensation of the lens, simultaneously acquire new bimodal perception data during the user's gaze switching process, and adjust the individual initial compensation coefficient matrix. A Perform online updates.

[0182] Specifically, after each refractive compensation of the lens, new bimodal perception data during the user's gaze switching process is simultaneously acquired, and the individual initial compensation coefficient matrix A is updated online. Specifically, after completing the lens adjustment in step S2300 or S2400, the updated head posture data, gaze distance, and ciliary electromyography (CEMG) signals are acquired in real time using the nine-axis inertial sensor and dry electrode bioelectric sensor built into the head-mounted device, and the new data is spatiotemporally aligned with historical data. After bandpass filtering and baseline correction, the aligned data is processed to extract the updated CEMG energy. Simultaneously, based on the discrete interval index of the current head pitch and yaw angles, the coefficient 'a' to be updated in the individual initial compensation coefficient matrix A is located. ijk,lThe online update employs a sliding time window mechanism, using the most recent N² refractive compensation operation data (e.g., N²=50 times) as the training set. An incremental learning algorithm is used to progressively correct the coefficients in matrix A. During incremental learning, the weight allocation between new and existing data is dynamically adjusted based on the data timestamp, with newer data receiving higher weights (e.g., a linear decay factor λ=0.95). This ensures that matrix A can adapt to changes in individual users' accommodation habits (e.g., head movement pattern shifts due to changes in the learning environment). This step addresses the mismatch between the initial compensation coefficient matrix A and real-time physiological states caused by long-term user use. For example, in children, the ciliary muscle's accommodation ability gradually increases during growth and development, and the initial modeling data cannot cover this dynamic change. Without this step, the compensation strategy will gradually deviate from the user's actual needs, leading to the accumulation of refractive compensation lag errors, ultimately causing an increase in the proportion of ineffective accommodation movements and exacerbating eye strain. This step, together with the multiple regression modeling in step S1233, forms an iterative optimization closed loop. Through an online update mechanism, the real-time acquired dual-modal sensing data is injected back into the initial compensation model, causing matrix A to evolve from static baseline data into a dynamically adapted model, thereby improving the long-term stability of the lens compensation strategy. Furthermore, this step interacts with the adjusted lag prediction model in step S2200. The updated matrix A provides the prediction model with the latest coefficient baseline, avoiding amplified prediction bias caused by outdated compensation coefficients. The synergy between these two steps makes the triggering timing of the advanced compensation acceleration mechanism more accurate in emergency events.

[0183] Example 3

[0184] This embodiment, based on Embodiments 1 and 2, provides a refractive compensation device based on dual-modal sensing, such as... Figure 6 As shown, it includes:

[0185] Partitioning module: Used to establish the coordinate system O-XYZ of the head-mounted device, and to define the first and second regions in the coordinate system O-XYZ;

[0186] Dual-modal data acquisition module: The user performs back-and-forth gaze switching between the first and second regions at different switching speeds, and the module collects dual-modal perception data during the user's gaze switching process; the dual-modality refers to physiological signal mode and spatial motion mode;

[0187] Adjustment event determination module: Based on bimodal perception data during user gaze switching, it determines emergency adjustment events and stable adjustment events;

[0188] Refractive compensation module: If an emergency accommodation event is identified, an advanced compensation acceleration mechanism is triggered; if a stable accommodation event is identified, a conventional compensation mechanism is used.

[0189] In addition, the parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0190] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A refractive compensation method based on bimodal sensing, characterized in that, The method includes: A coordinate system O-XYZ is established for the head-mounted device. A first region and a second region are defined in the coordinate system O-XYZ. The user performs a back-and-forth gaze switching between the first region and the second region at different switching speeds, and bimodal perception data is collected during the user's gaze switching process. The bimodality refers to physiological signal mode and spatial motion mode. Based on bimodal perception data during user gaze switching, emergency adjustment events and stable adjustment events are determined. If an emergency adjustment event is determined, an advance compensation acceleration mechanism is triggered; if a stable adjustment event is determined, a conventional compensation mechanism is used. The method for establishing the body coordinate system O-XYZ of the head-mounted device is as follows: the midpoint of the line connecting the user's pupils is taken as the origin O, the X-axis runs from left to right along the line connecting the pupils, the Y-axis points to the top of the head, and the Z-axis is defined as the direction of the line of sight directly in front. The method for defining the first and second regions in the coordinate system O-XYZ is as follows: the interval Z∈[a,b] is defined as the first region, and the interval Z∈[c,d] is defined as the second region, where a,b,c,d are positive real numbers and a<b<c<d; The dual-modal sensing data includes physiological signal data and spatial motion data; the physiological signal data refers to ciliary electromyography (CEMG) signals; the spatial motion data includes the user's head posture data, gaze distance, and three-dimensional gaze point coordinates. The method for determining emergency accommodation events and stable accommodation events is as follows: based on head posture data and gaze distance, the distance change rate V is calculated. D Head motion gradient G H And the three-dimensional trajectory curvature κ; calculate the slope ke of the change in the ciliary muscle electromyography signal; if and and and >k th If the event is positive, it is classified as an emergency adjustment event; otherwise, it is classified as a stable adjustment event. The preset distance change rate threshold, The preset head motion gradient threshold, k is the preset trajectory curvature threshold. th The preset slope threshold; The aforementioned advance compensation acceleration mechanism is: dynamically shortening and adjusting the delay parameter. The shortened compensation delay time is obtained. ,based on Calculate the target refractive power in advance ,according to Adjust the lens; the adjustment delay parameter It was acquired synchronously during the collection of bimodal perception data during the user's gaze switching process.

2. The refractive compensation method based on bimodal sensing according to claim 1, characterized in that, The calculated distance change rate V D The method is as follows: using a sliding time window, the distance change rate V is calculated based on the gaze distance. D .

3. The refractive compensation method based on bimodal sensing according to claim 2, characterized in that, The three-dimensional trajectory curvature The calculation method is as follows: A global space curve is generated from the coordinates of the 3D gaze point; the current gaze point is selected from the global space curve. Compared to the past One gaze point The curvature of the constructed local space curve is calculated as the curvature of the three-dimensional trajectory. .

4. A refractive compensation device based on bimodal sensing, used to implement the refractive compensation method based on bimodal sensing according to any one of claims 1-3, characterized in that, The device includes: Partitioning module: Used to establish the coordinate system O-XYZ of the head-mounted device, and to define the first and second regions in the coordinate system O-XYZ; Dual-modal data acquisition module: The user performs back-and-forth gaze switching between the first and second regions at different switching speeds, and the module collects dual-modal perception data during the user's gaze switching process; the dual-modality refers to physiological signal mode and spatial motion mode; Adjustment event determination module: Based on bimodal perception data during user gaze switching, it determines emergency adjustment events and stable adjustment events; Refractive compensation module: If an emergency accommodation event is identified, an advanced compensation acceleration mechanism is triggered; if a stable accommodation event is identified, a conventional compensation mechanism is used.

Images