A mobile terminal-based visual training scheme adaptive generation method and system

Abstract

The application discloses a kind of visual training scheme self-adaptive generation method and system based on mobile terminal, the method includes: obtaining user basic information;Constitute and include multiple fixation target targets fixation digital target set, based on the fixation digital target set, determine fixation limited point set;According to the fixation limited point set, determine the deviation state type;Obtain the fixation duration of each fixation target target of user, generate fixation maintenance parameter according to the fixation duration, and calculate fixation distribution deflection angle value;According to the deviation state type and the fixation distribution deflection angle value, and generate visual training scheme in combination with the user basic information.The method and system provided by the application can realize the individualization self-adaptive generation of visual training scheme.

Description

Technical Field

[0001] This invention relates to the field of intelligent applications for mobile terminals, and in particular to an adaptive generation method and system for visual training schemes based on mobile terminals. Background Technology

[0002] Binocular co-focus is a crucial component of visual function. When both eyes cannot simultaneously and stably maintain focus on the same target, a phenomenon known as gaze deviation may occur. Adaptive training targeting this gaze deviation is an important way to improve visual function and promote binocular coordination.

[0003] Currently, training schemes for gaze shift states mainly suffer from the following problems:

[0004] (1) Existing detection methods generally rely on professional optical equipment and fixed sites, resulting in high detection costs and complex operation procedures, which are difficult to meet the daily needs of households;

[0005] (2) There is a lack of portable self-testing tools for gaze function that are suitable for home scenarios, making it impossible to make early predictions and long-term continuous monitoring of the state of visual function.

[0006] (3) Existing training mostly adopts a general preset mode, which cannot be adapted to the specific type and degree of individual gaze shift, resulting in insufficient training targeting.

[0007] (4) The training intensity parameters are mostly fixed values ​​or set by human experience. There is no quantitative correlation mechanism between the detection results and the training intensity, and there is a lack of adaptive adjustment logic.

[0008] (5) The training process lacks dynamic adjustment capabilities and cannot automatically adapt the training stage and intensity based on the user's real-time training performance and improvement progress, which may easily lead to problems such as low training efficiency or imbalance in intensity matching.

[0009] (6) There is no intelligent system that can automatically generate and continuously iterate and optimize personalized training programs by integrating multiple dimensions of information such as age characteristics, gaze deviation direction, deviation degree and historical training effect.

[0010] Therefore, developing an adaptive generation method and system for visual training schemes based on mobile terminals to achieve convenient detection of gaze function in home scenarios and personalized adaptive generation of training schemes, while being compatible with both naked-eye and virtual reality (VR) usage modes, is a technical problem that urgently needs to be solved by technicians in this field. Summary of the Invention

[0011] In view of the above-mentioned shortcomings of current training schemes for gaze shift states, the present invention provides an adaptive generation method for visual training schemes based on mobile terminals, which can realize personalized adaptive generation of visual training schemes.

[0012] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0013] An adaptive generation method for visual training schemes based on mobile terminals, the method comprising:

[0014] Obtain basic user information;

[0015] Construct a set of gaze digital targets including multiple gaze targets; determine a set of gaze-limited points based on the set of gaze digital targets; and determine the type of offset state based on the set of gaze-limited points.

[0016] The user's gaze duration for each of the aforementioned gaze targets is obtained, gaze maintenance parameters are generated based on the gaze duration, and gaze distribution skewness values ​​are calculated.

[0017] A visual training scheme is generated based on the offset state type and the gaze distribution skew angle value, combined with the user's basic information.

[0018] According to one aspect of the present invention, the construction of a gaze digital target set including multiple gaze target targets, the determination of a gaze-limited point set based on the gaze digital target set, and the determination of the offset state type based on the gaze-limited point set specifically include:

[0019] A set of gaze-based digital targets, comprising multiple gaze targets, is constructed based on a standardized coordinate system.

[0020] In response to the user's selection of the gaze target, a set of gaze-limited points is determined, and the gaze-limited centroid vector is calculated;

[0021] The type of deviation state is determined based on the preset gaze deviation determination rules and the gaze-limited center of gravity vector.

[0022] According to one aspect of the present invention, the step of determining the set of gaze-limited points and calculating the gaze-limited centroid vector in response to the user's selection of the gaze target specifically includes:

[0023] Construct a set of gaze-restricted points :

[0024]

[0025] In the formula, Let be the two-dimensional coordinate vector of the i-th gaze-limited point; n is the total number of gaze-limited points.

[0026] right The mean value is applied to obtain the gaze-limited centroid vector. :

[0027]

[0028] In the formula, The number of restricted fixation points. for The x-axis component, for The ordinate components, , The two-dimensional coordinate vectors of the i-th gaze-limited point are respectively. The horizontal and vertical components of the coordinates.

[0029] According to one aspect of the present invention, the preset gaze deviation determination rule includes a first determination rule and a second determination rule, based on the gaze-limited centroid vector. The first determination rule is:

[0030] like It is determined to be a type I horizontal offset state;

[0031] like It is determined to be a second type of horizontal offset state;

[0032] like It is determined to be a type I vertical offset state;

[0033] like It is determined to be a type II vertical offset state;

[0034] in, The horizontal threshold, , For vertical threshold, R is the preset distribution radius.

[0035] According to one aspect of the present invention, the second determination rule is specifically as follows:

[0036] Invert the determination condition for the x-coordinate component of the gaze-restricted centroid vector in the first determination rule:

[0037] like It is determined to be a type I horizontal offset state;

[0038] like It is determined to be a second type of horizontal offset state;

[0039] The determination logic for the ordinate component of the gaze-limited centroid vector in the second determination rule is the same as that in the first determination rule.

[0040] According to one aspect of the present invention, the step of obtaining the user's gaze duration for each of the gaze targets, generating gaze maintenance parameters based on the gaze duration, and calculating the gaze distribution skew angle specifically includes:

[0041] The duration of the user's gaze is obtained based on the set of gaze digital optotypes;

[0042] The fixation success rate is calculated based on the fixation duration, and the fixation maintenance parameters are determined.

[0043] The gaze distribution skew angle value is obtained based on the gaze maintenance parameters.

[0044] According to one aspect of the present invention, the step of generating a visual training scheme based on the offset state type and the gaze distribution skew angle value, combined with the user's basic information, specifically includes:

[0045] Construct training feature vectors based on the aforementioned user basic information;

[0046] Generate a visual training scheme based on the offset state type;

[0047] The training intensity of the visual training scheme is determined based on the training feature vector and the gaze distribution skew angle value.

[0048] According to one aspect of the present invention, generating a visual training scheme based on the offset state type specifically involves: presetting a clock direction training scheme; and matching a training scheme from the clock direction training scheme based on the offset state type.

[0049] According to one aspect of the present invention, the visual training scheme adaptive generation method further includes: employing a three-stage dynamic intensity adjustment algorithm based on the training feature vector.

[0050] An adaptive generation system for visual training schemes based on mobile terminals, the system being applied to the aforementioned adaptive generation method for visual training schemes based on mobile terminals, the system comprising:

[0051] The data acquisition module retrieves basic user information.

[0052] The pre-processing module constructs a set of gaze digital targets including multiple gaze targets, determines a set of gaze-limited points based on the set of gaze digital targets, and determines the type of offset state based on the set of gaze-limited points.

[0053] The post-processing module acquires the duration of the user's gaze at each of the gaze targets, generates gaze maintenance parameters based on the gaze duration, and calculates the gaze distribution skew angle value.

[0054] The result generation module generates a visual training scheme based on the offset state type, the gaze distribution skew angle value, and the user's basic information.

[0055] The advantages of this invention are as follows: It constructs a set of gaze targets using a clock coordinate system, collects user gaze-limited points and calculates the gaze-limited centroid vector, and combines this with preset gaze offset judgment rules to achieve objective judgment of gaze offset direction; it calculates gaze maintenance parameters and gaze distribution skewness values ​​through a digital sequence gaze annotation gaze duration quantification method, achieving refined classification of gaze offset degree; and it matches 12 preset refined clock direction training schemes according to the offset state type, realizing adaptive generation from detection results to training schemes. This invention features a user-friendly detection interface and simple operation, supporting both naked-eye and VR modes, reducing detection costs; and its personalized automatic matching of training schemes requires no manual intervention, significantly improving the targeting of training. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments 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.

[0057] Figure 1 This is a schematic flowchart of the adaptive generation method for the visual training scheme described in this invention;

[0058] Figure 2 This is a schematic diagram of the adaptive generation system for the visual training scheme described in this invention;

[0059] Figure 3 This is the naked-eye mode main interface of the adaptive generation system for the visual training scheme described in this invention;

[0060] Figure 4 This is the right-eye direction detection guidance interface for the naked-eye mode of the adaptive generation system of the visual training scheme described in this invention.

[0061] Figure 5 This is the left-eye direction detection guidance interface for the naked-eye mode of the adaptive generation system of the visual training scheme described in this invention.

[0062] Figure 6 This is the right-eye severity detection interface of the naked-eye mode of the adaptive generation system of the visual training scheme described in this invention.

[0063] Figure 7 This is the left-eye severity detection interface of the naked-eye mode of the adaptive generation system of the visual training scheme described in this invention.

[0064] Figure 8 This is the main interface of the VR mode of the adaptive generation system for the visual training scheme described in this invention.

[0065] Figure 9 This is the right-eye direction detection guidance interface for the VR mode of the adaptive generation system of the visual training scheme described in this invention.

[0066] Figure 10 This is the left-eye direction detection guidance interface for the VR mode of the adaptive generation system of the visual training scheme described in this invention.

[0067] Figure 11 This is the right eye severity detection interface of the VR mode of the adaptive generation system for the visual training scheme described in this invention.

[0068] Figure 12 This is the left-eye severity detection interface for the VR mode of the adaptive generation system of the visual training scheme described in this invention.

[0069] Figure 13 This is the clock position direction detection interface of the adaptive generation system of the visual training scheme described in this invention;

[0070] Figure 14 This is the interface for the degree detection progress indicator of the adaptive generation system of the visual training scheme described in this invention;

[0071] Figure 15 The orientation test of the adaptive generation system for the visual training scheme described in this invention is in a state where no selection is made.

[0072] Figure 16 The orientation test selection state interface of the adaptive generation system for the visual training scheme described in this invention;

[0073] Figure 17 This invention provides a direction-testing interface for the adaptive generation system of the visual training scheme, without any degree testing capability.

[0074] Figure 18 This is the test result display interface of the adaptive generation system for the visual training scheme described in this invention. Detailed Implementation

[0075] 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.

[0076] Example 1

[0077] like Figure 1 As shown, an adaptive generation method for a visual training scheme based on a mobile terminal is disclosed. The mobile terminal includes a display screen, a processor, a memory, a VR (Virtual Reality) device, a touch input module, an audio module, and a front-facing camera. The display screen is used to present a visual target interface and a training interface, including but not limited to a first visual target interface and a second visual target interface; the processor uses an ARM architecture chip with a main frequency of not less than 1.8GHz; the memory includes RAM and ROM for storing program code and detection data; the touch input module is used to receive user touch operations; the audio module is used to play voice prompts; the VR device is used for VR visual training, supporting Google Cardboard or similar mobile VR boxes, or all-in-one VR devices such as Pico and Oculus. The front-facing camera is used for head tracking in VR mode. The method is executed according to the following steps:

[0078] Step S1: Obtain basic user information;

[0079] The specific steps for obtaining basic user information in step S1 are as follows:

[0080] Basic user information is obtained through manual entry or batch retrieval, specifically including:

[0081] (1) Age segmentation: such as four age groups: 3 years old, 4 years old, 5 years old, and 6 years old and above;

[0082] (2) Gender: Boy or Girl;

[0083] (3) Wearing glasses: Yes or No.

[0084] Step S2: Construct a set of gaze digital targets including multiple gaze targets; determine a set of gaze-limited points based on the set of gaze digital targets; determine the type of offset state based on the set of gaze-limited points.

[0085] Step S2 involves constructing a set of gaze digital targets, including multiple gaze target targets, determining a set of gaze-limited points based on the set of gaze digital targets, and determining the offset state type based on the set of gaze-limited points, specifically including:

[0086] Step S21: Construct a set of gaze digital targets based on a standardized coordinate system, including multiple gaze target targets.

[0087] (1) Establish a standardized coordinate system

[0088] Using the center of the mobile terminal display screen as the origin Establish a two-dimensional Cartesian coordinate system.

[0089] (2) Calculate the standard coordinates of the gaze target.

[0090] The formula for calculating the standard coordinates of each clock position is as follows:

[0091]

[0092] In the formula, k is the clock position number. R is the radius of the point distribution; the angle is measured clockwise from the 12 o'clock position (directly above), for example... hour, , ,have to That is, directly to the right.

[0093] Thus, the 12 clock points (i.e., the coordinates of the gaze target) constitute the gaze digital target set. , where each element Let be the two-dimensional coordinate vector of the k-th gaze target.

[0094] In actual use, the clock display interface features a pure black background with a yellow cartoon smiley face in the center as the target of attention. The smiley face has a diameter of 15% of the screen height. The 12 clock points (i.e., the target visual markers) are arranged in a standard clock face layout on the circumference, with a spacing of 30 degrees between the points and a diameter of 8% of the screen width. Voice prompts are displayed in a bubble to guide the user to focus on the central target.

[0095] Step S22: In response to the user's selection of the gaze target, determine the set of gaze-limited points and calculate the gaze-limited centroid vector.

[0096] In practical use, the clock position interface on the display screen flashes each clock position sequentially (flash frequency 1Hz, flash duration 500ms), and the user follows the instructions by moving their eyes to focus on each position. When the user finds it difficult to focus on a certain clock position, they can click on that position (i.e., the position where focus is limited) through the touch input module, and the memory records the selection result (clock position number k and clock position coordinates). The test was performed in the order of right eye first, then left eye. Specifically:

[0097] Collect a set of user-marked gaze-limited points. :

[0098]

[0099] In the formula, Let be the two-dimensional coordinate vector of the i-th gaze-limited point; n is the total number of gaze-limited points.

[0100] right The weighted average of the coordinates of each clock position is used to obtain the gaze-limited centroid vector. :

[0101]

[0102] In the formula, The number of restricted fixation points. for The x-axis component, for The ordinate components, , The two-dimensional coordinate vectors of the i-th gaze-limited point are respectively. The horizontal and vertical components of the coordinates.

[0103] Step S23: Determine the type of deviation state based on the preset gaze deviation determination rule and the gaze-limited centroid vector.

[0104] In practical use, the preset gaze deviation determination rule includes a first determination rule (i.e., the right eye strabismus determination rule) and a second determination rule (i.e., the left eye strabismus determination rule), based on the gaze-limited centroid vector. The first determination rule is:

[0105] like It was determined to be a Class I horizontal deviation state, namely esotropia (eyeball deviates towards the nose).

[0106] like It was determined to be a type II horizontal deviation state, namely exotropia (the eyeball deviates towards the temporal side).

[0107] like It was determined to be a type I vertical offset state, namely, upward tilt (eyeball deviating upwards).

[0108] like It was determined to be a type II vertical offset state, namely downward tilt (the eyeball deviates downward).

[0109] in, The horizontal threshold, , For vertical threshold, R is a preset distribution radius. In this embodiment, W and H represent the screen width and screen height of the mobile terminal display, respectively.

[0110] In practical use, the determination condition for the x-coordinate component of the gaze-restricted centroid vector in the first determination rule is reversed:

[0111] like It was determined to be a Class I horizontal deviation state, namely esotropia (eyeball deviates towards the nose).

[0112] like It was determined to be a type II horizontal deviation state, namely exotropia (the eyeball deviates towards the temporal side).

[0113] The determination logic for the ordinate component of the gaze-limited centroid vector in the second determination rule is the same as that in the first determination rule.

[0114] For the left eye, since the two eyes are mirror-symmetrical in the horizontal direction, the direction of horizontal component determination is opposite to that of the right eye, while the rule for vertical component determination is the same as that of the right eye.

[0115] Furthermore, this embodiment also includes calculating the confidence level of the gaze-limited points, the calculation formula of which is:

[0116]

[0117] In the formula, The standard deviation is obtained by calculating the azimuth angle of each gaze-limited point in the standard coordinate system. If the result is deemed reliable, the user will be prompted to retest.

[0118] Furthermore, this embodiment supports matching detection in both naked-eye mode and VR mode. In naked-eye mode, the user needs to manually cover the non-detection eye; in VR mode, virtual covering is automatically achieved through dual-screen rendering technology.

[0119] Step S3: Obtain the duration of the user's gaze at each of the aforementioned gaze targets, generate gaze maintenance parameters based on the gaze duration, and calculate the gaze distribution skew angle value;

[0120] Step S3, which involves obtaining the user's gaze duration for each of the stated gaze targets, generating gaze maintenance parameters based on the gaze duration, and calculating the gaze distribution skew angle, specifically includes:

[0121] In actual use, numerical targets (i.e., gaze targets) are displayed sequentially in the center of the screen interface, with numbers ranging from 1 to 12. The numerical targets are displayed one by one in ascending order, and each numerical target has a base display time. =3 seconds.

[0122] Step S31: Obtain the user's gaze duration based on the gaze digital optotype set.

[0123] In practical use, a high-precision timer (10ms accuracy) is started synchronously. The user continuously gazes at the currently displayed digital target. When they feel their gaze is restricted and cannot maintain a stable, clear gaze, they touch the screen to confirm "restricted gaze," and the memory records the duration of this gaze. .

[0124] Step S32: Calculate the fixation success rate based on the fixation duration and determine the fixation maintenance parameters.

[0125] The formula for calculating the fixation success rate of digital targets is:

[0126] )

[0127] In the formula, , It reflects the user's ability to maintain a stable gaze on the current digital target; 1.0 indicates that it can be maintained until the end of the base time.

[0128] Starting from a certain number k, three consecutive numbers If all values ​​are less than 0.5, the eye test is immediately terminated, and the position of the first failed value (i.e., the monocular fixation maintenance parameter) is taken: If the termination condition has not been triggered after testing the number 12, then... .therefore The larger the value, the stronger the sustained gaze ability.

[0129] Perform steps S31 and S32 on the right eye and left eye respectively to obtain... (i.e., the right eye) )and (i.e., the left eye) The smaller of the two values ​​is taken as the fixation maintenance parameter (i.e., the final strabismus score): .

[0130] Step S33: Obtain the gaze distribution skew angle value based on the gaze maintenance parameters.

[0131]

[0132] The unit of deviation is degrees, and its value varies with... Linear change.

[0133] Furthermore, the gaze deviation level is determined based on the gaze maintenance parameters.

[0134] according to Perform piecewise linear mapping to determine the fixation maintenance level (i.e., the strabismus degree level):

[0135] when This is a severe case;

[0136] when It is moderate;

[0137] when It is mild;

[0138] when , is tiny;

[0139] when No strabismus.

[0140] In practical use, the results of combining the strabismus state type and the gaze distribution deviation value are used to generate a detection report, including: strabismus type (esotropia / exotropia / no strabismus), strabismus degree (12 levels), detection time, and detection mode. The report is simultaneously stored in a local database and synchronized to a cloud server.

[0141] Step S4: Generate a visual training scheme based on the offset state type and the gaze distribution skew angle value, combined with the user basic information.

[0142] Step S4, which generates a visual training scheme based on the offset state type and the gaze distribution skew angle value, and in conjunction with the user's basic information, specifically includes:

[0143] Step S41: Construct training feature vectors based on the user basic information

[0144] The training feature vectors are constructed as follows:

[0145]

[0146] In the formula, The age correction factor is determined by the user's age segment. In this embodiment, the value is 0.5 for 3 years old, 0.65 for 4 years old, 0.8 for 5 years old, and 1.0 for 6 years old and above. This represents the historical training progress rate; the default value for the first test is 0.5. The complete calculation method is as follows:

[0147] (1) Define the quality score of a single training session :

[0148]

[0149] In the formula, the weighting coefficients , , ; Let be the fixation accuracy during the i-th training iteration. Let be the completion rate of the i-th training iteration. Let be the training duration target rate for the i-th training session.

[0150] (2) Calculate the historical training progress rate :

[0151]

[0152] In the formula, The average quality score of the user's three most recent effective training sessions. This is the average quality score of the user's three most recent effective training sessions.

[0153] Step S42: Generate a visual training scheme based on the offset state type.

[0154] There are 12 preset refined clock direction training schemes. Each scheme consists of training eye type, strabismus type, vertical offset direction, and three sets of training paths and number of circles (A / B / C), as shown in Table 1.

[0155] Table 1. 12 Refined Clock Direction Training Schemes

[0156]

[0157] It should be noted that in the "middle" column of the "eye-strabismus type-vertical direction" column in Table 1, "middle" indicates that there is no obvious vertical offset component and the training path is mainly in the horizontal direction.

[0158] Subsequently, a training scheme is matched from the clock direction training scheme based on the offset state type, and the matching results are as follows:

[0159] Internal oblique corresponding to external abduction training, program range: number 1, 2, 3, 10, 11, 12;

[0160] External oblique corresponding to adduction training, program range: number 4, 5, 6, 7, 8, 9;

[0161] Upward slope corresponds to downward pressure training; program range: numbers 3, 6, 9, 12;

[0162] Downsloping exercises correspond to upward lifting exercises. The training program ranges from numbers 1, 4, 7, and 10.

[0163] Finally, by combining the user's training eye (left / right) and vertical offset components (top / middle / bottom), the correct scheme number is accurately matched from the scheme range of the corresponding training type to generate the user's personalized visual training scheme.

[0164] Step S43: Based on the training feature vector and combined with the gaze offset angle value, determine the training intensity of the visual training scheme.

[0165] (1) The formula for calculating training intensity is:

[0166] Base strength: Normalization to Interval.

[0167] Overall strength: , .

[0168] (2) Training parameter output:

[0169] Duration of a single training session: ;

[0170] Number of round trips: In the formula, round is the rounding function.

[0171] Target movement speed: , This is the preset reference speed.

[0172] (3) Training cycle: It is recommended to train 1-2 times a day, based on the progress rate of 3 consecutive training sessions. Dynamically adjust training frequency and intensity, if Then appropriately increase the intensity, if the rate of progress Then reduce the intensity and recommend seeking medical attention. Then the current training intensity will remain unchanged.

[0173] Furthermore, the adaptive generation method for the visual training scheme also includes: employing a three-stage dynamic intensity adjustment algorithm based on the training feature vector, specifically:

[0174] The training employs a three-stage dynamic intensity adjustment algorithm, which automatically determines the timing of each stage entry based on historical training data.

[0175] (1) Adaptation period (initial stage): Training intensity coefficient target movement speed Dwell time at each clock position Starting from 1 second, press Linearly increasing to 6 seconds, This represents the total number of training iterations for this phase, where n indicates the number of training iterations the user has completed in the current phase. Phase upgrade condition: Average fixation success rate over 5 consecutive training sessions. And in the last 3 training sessions Compared to the initial Improve by 2 places or more.

[0176] (2) Improvement period: training intensity coefficient target movement speed , The time increments linearly to 8 seconds. Stage advancement condition: 5 consecutive training sessions. ,and Compared to the initial Improvement by 4 digits or more; Stage downgrade criteria: Average fixation success rate over 3 consecutive training sessions It will automatically revert to the adaptation period.

[0177] (3) Intensive period: training intensity coefficient target movement speed , The duration increases linearly to 10 seconds. The reinforcement phase can continue until... (Minor) or user-initiated termination, the total training cycle can be flexibly adapted to 4-12 weeks.

[0178] Among them, the average fixation success rate fixation success rate based on digital targets The calculation yielded the result.

[0179] Example 2

[0180] like Figure 2-18 As shown, an adaptive generation system for visual training schemes based on mobile terminals is described in this embodiment. This embodiment applies to the adaptive generation method for visual training schemes based on mobile terminals described in Embodiment 1. The system includes:

[0181] Data acquisition module M1 acquires basic user information;

[0182] The pre-processing module M2 constructs a set of gaze digital targets including multiple gaze targets, determines a set of gaze-restricted points based on the set of gaze digital targets, and determines the type of offset state based on the set of gaze-restricted points.

[0183] The post-data processing module M3 acquires the duration of the user's gaze at each of the gaze targets, generates gaze maintenance parameters based on the gaze duration, and calculates the gaze distribution skew angle value.

[0184] The result generation module M4 generates a visual training scheme based on the offset state type, the gaze distribution deviation angle value, and the user's basic information.

[0185] The advantages of this invention are as follows: It constructs a set of gaze targets using a clock coordinate system, collects user gaze-limited points and calculates the gaze-limited centroid vector, and combines this with preset gaze offset judgment rules to achieve objective judgment of gaze offset direction; it calculates gaze maintenance parameters and gaze distribution skewness values ​​through a digital sequence gaze annotation gaze duration quantification method, achieving refined classification of gaze offset degree; and it matches 12 preset refined clock direction training schemes according to the offset state type, realizing adaptive generation from detection results to training schemes. This invention features a user-friendly detection interface and simple operation, supporting both naked-eye and VR modes, reducing detection costs; and its personalized automatic matching of training schemes requires no manual intervention, significantly improving the targeting of training.

[0186] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An adaptive generation method for visual training schemes based on mobile terminals, characterized in that, The method includes: Obtain basic user information; Construct a set of gaze digital targets including multiple gaze targets; determine a set of gaze-limited points based on the set of gaze digital targets; and determine the type of offset state based on the set of gaze-limited points. The user's gaze duration for each of the aforementioned gaze targets is obtained, gaze maintenance parameters are generated based on the gaze duration, and gaze distribution skewness values ​​are calculated. A visual training scheme is generated based on the offset state type and the gaze distribution deviation angle value, combined with the user's basic information.

2. The adaptive generation method for visual training schemes according to claim 1, characterized in that, The construction of a gaze digital target set includes multiple gaze target visual targets; based on the gaze digital target set, a gaze-limited point set is determined; and the determination of the offset state type based on the gaze-limited point set specifically includes: A set of gaze-based digital targets, comprising multiple gaze targets, is constructed based on a standardized coordinate system. In response to the user's selection of the gaze target, a set of gaze-limited points is determined, and the gaze-limited centroid vector is calculated; The type of deviation state is determined based on the preset gaze deviation determination rules and the gaze-limited center of gravity vector.

3. The adaptive generation method for visual training schemes according to claim 2, characterized in that, The process of responding to the user's selection of the gaze target, determining the set of gaze-limited points, and calculating the gaze-limited centroid vector specifically includes: Construct a set of gaze-restricted points : In the formula, Let be the two-dimensional coordinate vector of the i-th gaze-limited point; n is the total number of gaze-limited points. right After averaging, the gaze-limited centroid vector is obtained. : In the formula, The number of restricted fixation points. for The x-axis component, for The ordinate components, , The two-dimensional coordinate vectors of the i-th gaze-limited point are respectively. The horizontal and vertical components of the coordinates.

4. The adaptive generation method for visual training schemes according to claim 3, characterized in that, The preset gaze deviation determination rule includes a first determination rule and a second determination rule, based on the gaze-limited centroid vector. The first determination rule is: like It is determined to be a type I horizontal offset state; like It is determined to be a type II horizontal offset state; like It is determined to be a type I vertical offset state; like It is determined to be a type II vertical offset state; in, The horizontal threshold, , For vertical threshold, R is the preset distribution radius.

5. The adaptive generation method for visual training schemes according to claim 4, characterized in that, The second determination rule is as follows: Invert the determination condition for the x-coordinate component of the gaze-restricted centroid vector in the first determination rule: like It is determined to be a type I horizontal offset state; like It is determined to be a type II horizontal offset state; The determination logic for the ordinate component of the gaze-limited centroid vector in the second determination rule is the same as that in the first determination rule.

6. The adaptive generation method for visual training schemes according to claim 5, characterized in that, The steps of obtaining the user's gaze duration for each of the stated gaze targets, generating gaze maintenance parameters based on the gaze duration, and calculating the gaze distribution skew angle specifically include: The duration of the user's gaze is obtained based on the set of gaze digital optotypes; The fixation success rate is calculated based on the fixation duration, and the fixation maintenance parameters are determined. The gaze distribution skew angle value is obtained based on the gaze maintenance parameters.

7. The adaptive generation method for visual training schemes according to claim 1, characterized in that, The step of generating a visual training scheme based on the offset state type and the gaze distribution skew angle value, combined with the user's basic information, specifically includes: Construct training feature vectors based on the aforementioned user basic information; Generate a visual training scheme based on the offset state type; The training intensity of the visual training scheme is determined based on the training feature vector and the gaze distribution skew angle value.

8. The adaptive generation method for visual training schemes according to claim 7, characterized in that, The step of generating a visual training scheme based on the offset state type specifically involves: pre-setting a clock direction training scheme; and matching a training scheme from the clock direction training scheme based on the offset state type.

9. The adaptive generation method for visual training schemes according to claim 7, characterized in that, The adaptive generation method for the visual training scheme further includes: using a three-stage dynamic intensity adjustment algorithm based on the training feature vector.

10. An adaptive generation system for visual training schemes based on mobile terminals, characterized in that, The system is applied to the adaptive generation method for a visual training scheme based on a mobile terminal as described in any one of claims 1 to 9, and the system comprises: The data acquisition module retrieves basic user information. The pre-processing module constructs a set of gaze digital targets including multiple gaze targets, determines a set of gaze-limited points based on the set of gaze digital targets, and determines the type of offset state based on the set of gaze-limited points. The post-processing module acquires the duration of the user's gaze at each of the gaze targets, generates gaze maintenance parameters based on the gaze duration, and calculates the gaze distribution skew angle value. The result generation module generates a visual training scheme based on the offset state type, the gaze distribution skew angle value, and the user's basic information.

Images