Dynamic sighting mark generation and gesture feedback interactive vision test system

Abstract

The invention relates to the technical field of vision testing, in particular to a dynamic sighting mark generation and gesture feedback interactive vision testing system which comprises a sighting mark generation module, a gesture recognition module and a feedback adjustment module. The sighting mark generation module generates an initial sighting mark according to a preset rule and dynamically adjusts the size, position and shape of the initial sighting mark; the gesture recognition module captures gesture actions through a camera and analyzes direction, speed and force information; and the feedback adjustment module calculates sighting mark adjustment parameters based on the gesture analysis result, and optimizes a display strategy in combination with ambient light and screen brightness. Through non-contact gesture interaction and dynamic sighting mark adjustment, the accuracy and the individuation degree of the test result are improved, predictive adjustment is achieved based on deep learning, and the user experience and the intelligent level are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of medical testing and human-computer interaction technology, specifically an interactive vision testing system with dynamic optotype generation and gesture feedback. Background Technology

[0002] Visual acuity testing is a crucial means of assessing human visual function, widely used in ophthalmology, vision screening, and daily health management. Traditional visual acuity testing methods typically rely on static optotype charts (such as the Snellen chart or the E chart) and are conducted through manual instruction or verbal feedback. However, these methods have certain limitations: First, the fixed pattern of static optotypes can easily lead to memory effects on test results, reducing accuracy and objectivity; second, traditional testing requires manual recording and evaluation of results, which is inefficient and prone to human error; furthermore, most existing visual acuity testing systems lack interactivity and struggle to adapt to the personalized needs of different users, especially children or special populations, resulting in poor user experience and low test compliance.

[0003] In recent years, with the development of computer vision and human-computer interaction technologies, some vision testing systems based on dynamic optotypes have been proposed. These systems can reduce the impact of memory effects on test results to some extent by generating randomized or dynamically changing optotype patterns. However, existing technologies still have shortcomings: on the one hand, the algorithms for generating dynamic optotypes are often relatively simple, making it difficult to achieve highly randomized and diverse optotype presentations, thus limiting the flexibility and comprehensiveness of the test; on the other hand, existing interaction methods mostly rely on traditional button or touchscreen operations, which cannot fully simulate natural human behavior, especially lacking efficient and intuitive solutions in gesture recognition and feedback, resulting in insufficient user engagement and a less smooth testing process.

[0004] Therefore, developing an interactive vision testing system that combines dynamic optotype generation with natural gesture feedback can not only improve the accuracy and efficiency of the test, but also significantly improve the user experience and meet the needs of diverse testing scenarios, which has important research value and practical application significance. Summary of the Invention

[0005] The technical problem solved by this invention is that vision testing systems in related technologies lack dynamism and interactivity, and cannot adjust the test targets according to real-time user feedback, resulting in low accuracy and personalization of test results. At the same time, they fail to combine gesture operation to achieve contactless interaction, resulting in a monotonous and unintelligent user experience.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an interactive vision testing system with dynamic optotype generation and gesture feedback, comprising an optotype generation module, a gesture recognition module, and a feedback adjustment module;

[0007] The target generation module is used to generate initial targets according to preset rules, and dynamically adjust the size, position and shape of the targets according to the user's gesture feedback;

[0008] The gesture recognition module captures the user's hand gestures through a camera, analyzes the direction, speed, and force information of the gestures, and sends the analysis results to the feedback adjustment module.

[0009] The feedback adjustment module calculates the target adjustment parameters based on the gesture parsing results, sets the target change threshold, compares the new target parameters with the target change threshold, performs the first adjustment operation based on the comparison result, and establishes the first mapping relationship.

[0010] As a preferred embodiment of the interactive vision testing system with dynamic visual target generation and gesture feedback described in this invention, the initial visual target includes a basic graphic, color contrast, and display time. The basic graphic includes a circle, a square, and a triangle. The color contrast represents the brightness difference between the background color and the visual target color.

[0011] The gestures include swiping with one hand, zooming with both hands, and tapping with a fist. The gesture direction is indicated as horizontal or vertical. The gesture speed is indicated as the distance the gesture moves per unit time. The gesture force is indicated as the acceleration of the gesture.

[0012] The target generation module generates targets at intervals of 0.5 to 3 seconds, and the range of target changes covers the central and edge areas of the user's field of vision.

[0013] As a preferred embodiment of the interactive vision testing system with dynamic optotype generation and gesture feedback described in this invention, the calculation logic for the optotype adjustment parameters includes:

[0014] Obtain the direction and speed of two adjacent gestures, calculate the difference between adjacent gestures and record it as the first difference, calculate the difference in the corresponding gesture force and record it as the second difference, normalize the first difference and the second difference to obtain normalized values, and map the normalized values ​​to the adjustment parameters of the target size, position and shape respectively. When the gesture direction is horizontal, the position of the target is adjusted first, and when the gesture direction is vertical, the size of the target is adjusted first.

[0015] As a preferred embodiment of the interactive vision testing system with dynamic optotype generation and gesture feedback described in this invention, the optotype change category represents the category of the degree of change of the current state of the optotype relative to the previous state. The change category includes a first category, a second category, a third category, and a fourth category. The first category and the second category represent changes in optotype shrinkage or positional shift. The change magnitude of the first category is greater than that of the second category. The third category represents changes in optotype magnification or unchanged shape. The fourth category represents changes in optotype shape with a change magnitude greater than that of the third category.

[0016] As a preferred embodiment of the interactive vision testing system with dynamic optotype generation and gesture feedback described in this invention, the classification logic for the adjustment time corresponding to each optotype change category includes:

[0017] The optotype adjustment parameters are cleaned, and the corresponding gesture action time points are converted into time periods at the second level. The number of clusters is set to 4, and the optotype adjustment parameters are classified into various change categories. Each optotype adjustment parameter is labeled with a change category. The labeled optotype adjustment parameters are input into the K-means algorithm to obtain the time period corresponding to each change category. The time period is represented as A seconds to B seconds, where the value of A is less than B, and both A and B are distributed between 0 and 60. The time period corresponding to the fourth category is set as the adjustment time, which represents the key adjustment time within the time period of the gesture action.

[0018] As a preferred embodiment of the interactive vision testing system with dynamic visual target generation and gesture feedback described in this invention, the system acquires the ambient light intensity and device screen brightness of each test user within a first historical time period, and acquires the key adjustment time corresponding to each test user.

[0019] Set the first value to the ideal ambient light intensity and the second value to the ideal screen brightness. Obtain the key adjustment time corresponding to the ideal ambient light intensity and ideal screen brightness, and record it as the ideal adjustment time. Calculate the difference between the ambient light intensity and screen brightness of each test user and the first and second values, and record them as the light difference and brightness difference. Calculate the difference between the corresponding key adjustment time and ideal adjustment time, and record it as the time difference.

[0020] The weights of the environmental data are calculated, and the calculation logic for these weights includes:

[0021] Using light difference and brightness difference as independent variables and time difference as dependent variable, a multiple regression analysis was performed to obtain the coefficients of light difference and brightness difference. The coefficients of light difference and brightness difference were then set as the weights of ambient light intensity and screen brightness, respectively.

[0022] As a preferred embodiment of the interactive vision testing system with dynamic optotype generation and gesture feedback described in this invention, the following steps are taken: an interference function is constructed based on the weights, the interference function is used to calculate the deviation value of optotype adjustment time under the current environmental conditions, the current ambient light intensity and screen brightness are input into the interference model to obtain the current time deviation value, the current test user is obtained, the key adjustment time corresponding to the current test user is obtained, the current time deviation value and the key adjustment time are summed to obtain a first sum value, and the first sum value is set as the new key adjustment time.

[0023] As a preferred embodiment of the interactive vision testing system with dynamic optotype generation and gesture feedback described in this invention, the modeling logic of the feedback adjustment module includes:

[0024] The system acquires the user's historical gesture actions and corresponding target adjustment records, performs missing value processing and outlier processing on the historical gesture actions and target adjustment records, divides the user's field of view into partitions of a first size, the size of which is larger than the minimum display size of the target, and only one target can be displayed in each partition, acquires the position of the target after partitioning, and extracts the position matrix, size matrix and shape matrix respectively using the target position, target size and target shape as variables;

[0025] Set a sampling time step. If a target exists within the sampling time step, set the element corresponding to the position matrix to 1; otherwise, set the element corresponding to the position matrix to 0. Normalize the target size so that the target size is distributed between 0 and 1, and assign the normalized target size to the element corresponding to the size matrix. When the target shape is circular, set the element corresponding to the shape matrix of all partitions in the user's field of view to 1; otherwise, set the element corresponding to the shape matrix of all partitions in the user's field of view to 0.

[0026] After encoding using DTSE, the position matrix, size matrix, and shape matrix are encoded. A deep learning model is selected, and the encoded matrices are used as inputs to the model. The corresponding key adjustment time and the corresponding target change category are used as outputs to construct a feedback adjustment model.

[0027] As a preferred embodiment of the interactive vision testing system with dynamic optotype generation and gesture feedback described in this invention, the following steps are taken: The corresponding historical gesture actions and corresponding optotype adjustment records are input into the feedback adjustment model to obtain the key adjustment time and optotype change category; the current ambient light intensity and screen brightness are obtained; a new key adjustment time and a new optotype change category are obtained through an interference function; a third value is set as the optotype change threshold; the absolute value of the new optotype change category is compared with the optotype change threshold; and a first adjustment operation is performed based on the comparison result. The first adjustment operation includes changing the optotype display time and sending a completion signal.

[0028] When the absolute value of the new target change category is greater than or equal to the target change threshold, the first adjustment operation is set to change the target display time. The change includes continuously shortening or continuously extending the time until the absolute value of the new target change category is less than the target change threshold. Then, the change of the target display time is stopped and a completion signal is sent. When the absolute value of the new target change category is less than the target change threshold, the first adjustment operation is set to send a completion signal.

[0029] As a preferred embodiment of the interactive vision testing system with dynamic optotype generation and gesture feedback described in this invention, a first mapping relationship is established, which represents the mapping relationship between environmental data, the new key adjustment time of any test user, and the optotype display time. By inputting environmental data and the key adjustment time of any test user, the corresponding optotype display time is obtained.

[0030] The beneficial effects of this invention are as follows: By dynamically generating visual targets and combining them with gesture feedback to achieve contactless interaction, the system can adjust test conditions according to the user's real-time reactions, improving the accuracy and personalization of test results. At the same time, by combining environmental data such as ambient light and screen brightness, the system optimizes the visual target display strategy, enhancing the intelligent level of user experience. The deep learning-based modeling method enables the system to have predictive capabilities, allowing it to adjust visual target parameters in advance to adapt to the test needs of different users. The application of Internet of Things (IoT) technology enables the vision testing system to achieve remote monitoring and centralized management. Attached Figure Description

[0031] Figure 1 This is an overall structural block diagram of an interactive vision testing system with dynamic optotype generation and gesture feedback provided in an embodiment of the present invention.

[0032] Figure 2 This is a schematic diagram of the workflow of a target generation module provided in one embodiment of the present invention, including the generation logic of basic graphics, color contrast, and display time.

[0033] Figure 3 The flowchart of the gesture parsing module provided in one embodiment of the present invention illustrates the process of capturing and processing gesture direction, speed and force information.

[0034] Figure 4 This is a schematic diagram of the modeling logic of a feedback adjustment module provided in one embodiment of the present invention, including the construction and encoding process of the position matrix, size matrix and shape matrix.

[0035] Figure 5 The flowchart for establishing a first mapping relationship provided in one embodiment of the present invention illustrates the mapping logic between environmental data, key adjustment time, and target display time. Detailed Implementation

[0036] This invention provides an interactive vision testing system with dynamic optotype generation and gesture feedback. Its core lies in the collaborative work of an optotype generation module, a gesture recognition module, and a feedback adjustment module to achieve dynamic and interactive vision testing. The following is in conjunction with the appendix... Figure 1 To be continued Figure 5 The specific embodiments of the present invention will be described in detail below.

[0037] like Figure 1 As shown, the overall structure of this system includes a target generation module 101, a gesture recognition module 102, and a feedback adjustment module 103. The target generation module 101 is responsible for generating initial targets according to preset rules and dynamically adjusting the size, position, and shape of the targets based on user gesture feedback. The gesture recognition module 102 captures the user's gestures using a camera and analyzes their direction, speed, and force information. The feedback adjustment module 103 calculates target adjustment parameters based on the gesture analysis results, sets target change thresholds, performs a first adjustment operation based on the comparison results, and establishes a first mapping relationship. The cooperation between these modules enables the system to respond to user actions in real time, improving the accuracy and personalization of the test.

[0038] In practical implementation, the workflow of the target generation module 101 is as follows: Figure 2 As shown. First, the module generates an initial target, which includes a basic graphic, color contrast, and display time. The basic graphic can be a circle, square, or triangle, and the color contrast represents the brightness difference between the background color and the target color. The time interval for target generation is set to 0.5 to 3 seconds to ensure that the range of target changes covers the central and edge areas of the user's field of vision. For example, in a specific test scenario, when the user is in a low-light environment, the module will prioritize generating a high-contrast target to improve recognition. In addition, the target generation module 101 can also adjust the target's attributes based on the user's gesture feedback. For example, when the user makes a single-hand swipe gesture, the module will prioritize adjusting the target's position; while when the user makes a two-hand zoom gesture, the module will prioritize adjusting the target's size.

[0039] The working principle of gesture recognition module 102 is as follows: Figure 3As shown, this module captures the user's hand gestures via a camera and analyzes the direction, speed, and force of the gestures. Specifically, the gesture direction is divided into horizontal and vertical directions, the gesture speed is represented by the distance the gesture moves per unit time, and the gesture force is represented by the acceleration of the gesture. For example, when the user performs a single-hand swipe operation, the module records the horizontal direction and speed value of the gesture; when the user performs a two-hand zoom operation, the module analyzes the change in the force of the gesture. These analysis results are then sent to the feedback adjustment module 103 to calculate the target adjustment parameters.

[0040] The core function of the feedback adjustment module 103 is to calculate the target adjustment parameters based on the gesture analysis results and set the target change threshold. For example... Figure 4 As shown, the module first acquires the direction and speed of two adjacent gestures, calculates the difference between the adjacent gestures (referred to as the first difference), and simultaneously calculates the difference in the corresponding gesture strength (referred to as the second difference). After normalizing the first and second differences, normalized values ​​are obtained and mapped to the adjustment parameters for the target size, position, and shape. For example, when the user's gesture direction is horizontal, the module prioritizes adjusting the target position; when the gesture direction is vertical, the module prioritizes adjusting the target size. Furthermore, the module categorizes target changes into four types: the first and second types represent changes in target shrinkage or positional shift, with the first type showing a greater magnitude of change than the second; the third type represents changes in target enlargement or unchanged shape; and the fourth type represents changes in target shape with a greater magnitude than the third type. The adjustment times corresponding to these change categories are classified using the K-means algorithm. The specific logic is as follows: The optotype adjustment parameters are cleaned, the corresponding gesture action timestamps are converted into time intervals at the second level, the number of clusters is set to 4, the optotype adjustment parameters are categorized into each change category, and a label is assigned to each optotype adjustment parameter using the change category. The labeled optotype adjustment parameters are then input into the K-means algorithm to obtain the time intervals corresponding to each change category. For example, the time interval corresponding to the fourth category might be from A seconds to B seconds, where A is less than B and is evenly distributed between 0 and 60 seconds. This time interval is set as the critical adjustment time for subsequent optotype adjustments.

[0041] To further optimize the target adjustment strategy, the feedback adjustment module 103 also incorporates the impact of environmental data. For example... Figure 5As shown, the module first obtains the ambient light intensity and device screen brightness of each test user within the first historical time period, and also obtains the corresponding critical adjustment time for each test user. Then, it sets the first value as the ideal ambient light intensity and the second value as the ideal screen brightness, obtaining the critical adjustment time corresponding to the ideal ambient light intensity and ideal screen brightness, denoted as the ideal adjustment time. Next, it calculates the differences between the ambient light intensity and screen brightness of each test user and the first and second values, denoted as the light difference and brightness difference, and calculates the difference between the corresponding critical adjustment time and ideal adjustment time, denoted as the time difference. Through multiple regression analysis, using the light difference and brightness difference as independent variables and the time difference as the dependent variable, it obtains the coefficients of the light difference and brightness difference, setting these two coefficients as the weights of the ambient light intensity and screen brightness. For example, assuming the coefficient of the light difference is 0.6 and the coefficient of the brightness difference is 0.4, then the weight of the ambient light intensity is 0.6 and the weight of the screen brightness is 0.4. Based on these weights, the module constructs an interference function to calculate the deviation value of the target adjustment time under the current environmental conditions. The specific formula of the interference function is as follows:

[0042] ΔT=WL·ΔL+WB·ΔBΔT=WL·ΔL+WB·ΔB

[0043] Where ΔT represents the current time deviation, WL represents the weight of ambient light intensity, ΔL represents the light difference, WB represents the weight of screen brightness, and ΔB represents the brightness difference. By substituting the current ambient light intensity and screen brightness into the interference function, the module obtains the current time deviation. Subsequently, the module acquires the critical adjustment time corresponding to the current test user, sums the current time deviation value with the critical adjustment time, obtains the first sum, and sets this sum as the new critical adjustment time. For example, if the current critical adjustment time is 10 seconds and the current time deviation is 2 seconds, then the new critical adjustment time is 12 seconds.

[0044] The modeling logic of the feedback adjustment module 103 also includes the application of a deep learning model. The module first acquires the user's historical gestures and corresponding target adjustment records, and handles missing and outlier values. Then, it divides the user's field of view into partitions of a first size, where the partition size is larger than the minimum display size of the target, and only one target can be displayed in each partition. After obtaining the partitioned target positions, the module extracts the position matrix, size matrix, and shape matrix using target position, target size, and target shape as variables. Specifically, when a target exists within a sampling time step, the element corresponding to the position matrix is ​​set to 1; otherwise, it is set to 0. The target size is normalized to a range of 0 to 1, and the normalized target size is assigned to the element in the size matrix. When the target shape is circular, all partitions in the user's field of view are set to the element in the shape matrix; otherwise, they are set to 0. After DTSE encoding, the position matrix, size matrix, and shape matrix are encoded. The module selects a deep learning model, using the encoded matrices as inputs and the corresponding key adjustment time and target change category as outputs to construct a feedback adjustment model. For example, when the user's historical gesture is a single-handed swipe and the target adjustment record is a positional shift, the model can predict the future key adjustment time and target change category.

[0045] In practical applications, the workflow of the feedback adjustment module 103 is as follows: First, the corresponding historical gesture action and corresponding target adjustment record are input into the feedback adjustment model to obtain the key adjustment time and target change category. Then, the current ambient light intensity and screen brightness are obtained, and a new key adjustment time and a new target change category are obtained through an interference function. Next, the module sets a third value as the target change threshold and compares the absolute value of the new target change category with the target change threshold. If the absolute value of the new target change category is greater than or equal to the target change threshold, the module sets the first adjustment operation to change the target display time. The specific operation includes continuously shortening or continuously extending the target display time until the absolute value of the new target change category is less than the target change threshold, at which point the change in target display time stops, and a completion signal is sent. If the absolute value of the new target change category is less than the target change threshold, the module sets the first adjustment operation to send a completion signal. Finally, the module establishes a first mapping relationship, which represents the mapping relationship between environmental data, the new key adjustment time of any test user, and the target display time. By inputting environmental data and the key adjustment time of any test user, the corresponding target display time is obtained. For example, when the ambient light intensity is 500 lux and the screen brightness is 300 cd / m² 2 At that time, the module can quickly determine the target display time as 15 seconds based on the mapping relationship.

[0046] In summary, this invention achieves dynamic and interactive vision testing through the synergistic effect of the optotype generation module 101, gesture recognition module 102, and feedback adjustment module 103. The system not only adjusts testing conditions based on real-time user responses but also optimizes the optotype display strategy by incorporating environmental data such as ambient light and screen brightness. The deep learning-based modeling method enables the system to predictively adjust optotype parameters in advance to adapt to the testing needs of different users. The application of IoT technology allows for remote monitoring and centralized management of the vision testing system, enhancing the intelligence level of the user experience.

[0047] The foregoing description illustrates and describes several preferred embodiments of the present invention. However, as previously stated, it should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. An interactive vision testing system with dynamic optotype generation and gesture feedback, characterized in that, It includes a target generation module (101), a gesture recognition module (102), and a feedback adjustment module (103); the target generation module (101) is used to generate an initial target according to a preset rule, and dynamically adjust the size, position, and shape of the target according to the user's gesture feedback; The gesture recognition module (102) captures the user's gestures through the camera, analyzes the direction, speed and force information of the gestures, and sends the analysis results to the feedback adjustment module (103). The feedback adjustment module (103) calculates the target adjustment parameters based on the gesture analysis results, sets the target change threshold, compares the new target parameters with the target change threshold, performs the first adjustment operation according to the comparison result, and establishes the first mapping relationship.

2. The interactive vision testing system with dynamic optotype generation and gesture feedback as described in claim 1, characterized in that: The initial target includes a basic graphic, color contrast, and display time. The basic graphic includes a circle, a square, and a triangle. The color contrast is represented by the brightness difference between the background color and the target color. The gestures include swiping with one hand, zooming with both hands, and tapping with a fist. The gesture direction is indicated as horizontal or vertical. The gesture speed is indicated as the distance the gesture moves per unit time. The gesture force is indicated as the acceleration of the gesture. The target generation module (101) generates targets at intervals of 0.5 to 3 seconds, and the range of target changes covers the central and edge areas of the user's field of vision.

3. The interactive vision testing system with dynamic optotype generation and gesture feedback as described in claim 1, characterized in that: The calculation logic for the target adjustment parameters includes: Obtain the direction and speed of two adjacent gestures, calculate the difference between adjacent gestures and record it as the first difference, calculate the difference in the corresponding gesture force and record it as the second difference, normalize the first difference and the second difference to obtain normalized values, and map the normalized values ​​to the adjustment parameters of the target size, position and shape respectively. When the gesture direction is horizontal, the position of the target is adjusted first, and when the gesture direction is vertical, the size of the target is adjusted first.

4. The interactive vision testing system with dynamic optotype generation and gesture feedback as described in claim 1, characterized in that: The target change category represents the degree of change of the target's current state relative to the previous state. The change category includes a first category, a second category, a third category, and a fourth category. The first and second categories represent changes in target shrinkage or positional shift. The change magnitude of the first category is greater than that of the second category. The third category represents changes in target magnification or unchanged shape. The fourth category represents changes in target shape with a change magnitude greater than that of the third category.

5. The interactive vision testing system with dynamic optotype generation and gesture feedback as described in claim 1, characterized in that: The classification logic for adjustment time corresponding to each visual target change category includes: The optotype adjustment parameters are cleaned, and the corresponding gesture action time points are converted into time periods at the second level. The number of clusters is set to 4, and the optotype adjustment parameters are classified into various change categories. Each optotype adjustment parameter is labeled with a change category. The labeled optotype adjustment parameters are input into the K-means algorithm to obtain the time period corresponding to each change category. The time period is represented as A seconds to B seconds, where the value of A is less than B, and both A and B are distributed between 0 and 60. The time period corresponding to the fourth category is set as the adjustment time, which represents the key adjustment time within the time period of the gesture action.

6. The interactive vision testing system with dynamic optotype generation and gesture feedback as described in claim 1, characterized in that: Obtain the ambient light intensity and device screen brightness of each test user within the first historical time period, and obtain the corresponding key adjustment time for each test user; Set the first value to the ideal ambient light intensity and the second value to the ideal screen brightness. Obtain the key adjustment time corresponding to the ideal ambient light intensity and ideal screen brightness, and record it as the ideal adjustment time. Calculate the difference between the ambient light intensity and screen brightness of each test user and the first and second values, and record them as the light difference and brightness difference. Calculate the difference between the corresponding key adjustment time and ideal adjustment time, and record it as the time difference. The weights of the environmental data are calculated, and the calculation logic for these weights includes: Using light difference and brightness difference as independent variables and time difference as dependent variable, a multiple regression analysis was performed to obtain the coefficients of light difference and brightness difference. The coefficients of light difference and brightness difference were then set as the weights of ambient light intensity and screen brightness, respectively.

7. The interactive vision testing system with dynamic optotype generation and gesture feedback as described in claim 1, characterized in that: An interference function is constructed based on the weights. The interference function is used to calculate the deviation value of the target adjustment time under the current environmental conditions. The current ambient light intensity and screen brightness are input into the interference model to obtain the current time deviation value. The current test user is obtained, and the key adjustment time corresponding to the current test user is obtained. The current time deviation value and the key adjustment time are summed to obtain the first sum value. The first sum value is set as the new key adjustment time.

8. The interactive vision testing system with dynamic optotype generation and gesture feedback as described in claim 1, characterized in that: The modeling logic of the feedback adjustment module (103) includes: The system acquires the user's historical gesture actions and corresponding target adjustment records, performs missing value processing and outlier processing on the historical gesture actions and target adjustment records, divides the user's field of view into partitions of a first size, the size of which is larger than the minimum display size of the target, and only one target can be displayed in each partition, acquires the position of the target after partitioning, and extracts the position matrix, size matrix and shape matrix respectively using the target position, target size and target shape as variables; Set a sampling time step. If a target exists within the sampling time step, set the element corresponding to the position matrix to 1; otherwise, set the element corresponding to the position matrix to 0. Normalize the target size so that the target size is distributed between 0 and 1, and assign the normalized target size to the element corresponding to the size matrix. When the target shape is circular, set the element corresponding to the shape matrix of all partitions in the user's field of view to 1; otherwise, set the element corresponding to the shape matrix of all partitions in the user's field of view to 0. After encoding using DTSE, the position matrix, size matrix, and shape matrix are encoded. A deep learning model is selected, and the encoded matrices are used as inputs to the model. The corresponding key adjustment time and the corresponding target change category are used as outputs to construct a feedback adjustment model.

Images