Intelligent vision testing method and intelligent vision testing system

Abstract

The invention relates to the technical field of vision detection, in particular to an intelligent vision detection method and an intelligent vision test system. The intelligent vision detection method comprises the following steps: dynamically calculating a target visual angle and a size of a sighting target according to a test distance; generating the sighting mark by adopting an equivalent vision method; and displaying the sighting marks on a display device in a consistent manner at the target visual angle. The vision testee can independently complete vision detection under the condition that the vision testee is not limited by spatial distance, and the accuracy of a detection result is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of vision detection, and in particular relates to an intelligent vision detection method and an intelligent vision test system. Background Art

[0002] Currently, there are mainly two vision detection methods on the market. One is to use a traditional light box vision chart, and an operator conducts vision tests on the testee by pointing at the optotypes. The other is to display the optotypes through an electronic display device such as a monitor, and an operator conducts vision tests on the testee through a remote control or the like. When observing optotypes "E" or "C" in the above vision detection, the required spacing distance needs to meet the 5-meter distance required by GB / T 11533-2011. Therefore, the commonly used distances for current vision detection are 5m or 2.5m, and some electronic vision charts can reach 1m, all of which use the actual physical distance (5m) or a reflecting mirror (2.5m) or an equivalent method (the closest is 1m).

[0003] It can be seen that the existing vision detection methods have high requirements for spatial distance and cannot perform vision detection in a small space or on a desktop. Secondly, the existing vision detection methods generally require an operator to assist in completing the vision detection, and the vision testee cannot complete it independently, wasting manpower. Therefore, how to achieve vision detection that is not restricted by spatial distance, enabling the vision testee to complete the vision detection independently and improving the accuracy of the detection results has become an urgent problem to be solved. Summary of the Invention

[0004] In view of this, the present invention aims to provide an intelligent vision detection method and an intelligent vision test system, enabling the vision testee to complete the vision detection independently without being restricted by spatial distance and improving the accuracy of the detection results.

[0005] To achieve the above object, the technical solution of the present invention is realized as follows: The present invention provides an intelligent vision detection method, including the following steps: dynamically calculating the target viewing angle and size of the optotype according to the test distance; generating the optotype by using the equivalent vision method; and displaying the optotype on a display device with the target viewing angle being kept consistent.

[0006] Further, dynamically calculating the target viewing angle and size of the optotype according to the test distance specifically includes: dynamically calculating the viewing angle of the optotype according to the viewing angle formula; selecting the optimal one from the viewing angles of the optotype as the target viewing angle; and determining the size of the optotype based on the target viewing angle.

[0007] Further, generating the optotype by using the equivalent vision method specifically includes: calculating the equivalent viewing angle of the optotype based on the size of the optotype: (1) Wherein, is the equivalent viewing angle (in radians), is the pixel size of the visual target on the display device (in pixels), is the test distance (in centimeters), is the pixel density of the display device (in pixels per inch), and 2.54 is the conversion factor for converting inches to centimeters. Among them, the pixel size of the visual target on the display device : (2) The pixel density PPI of the display device: (3) Wherein, is the actual physical size of the visual target (in centimeters), is the resolution width of the display device, is the resolution height of the display device, is the diagonal size of the display device (in inches).

[0008] Further, after displaying the visual target on the display device with the target viewing angle being consistent, the method further includes: receiving and storing the behavioral data and the corresponding visual target information made by the tester, so as to output the prediction parameters of the visual target based on the behavioral data and the visual target information during the next test.

[0009] Further, displaying the visual target on the display device with the target viewing angle being consistent specifically includes: if the cumulative number of times of the tester's incorrect judgment of the viewing direction of the visual target reaches the set threshold, then display the size of the visual target that was correctly judged last time, and reset the number of incorrect times.

[0010] Further, before receiving and storing the behavioral data and the corresponding visual target information made by the tester, the method further includes: if the tester is a historical tester, dynamically adjusting the initial visual target parameters of the visual target based on the long short-term memory network LSTM model.

[0011] Further, if the tester is a historical tester, dynamically adjusting the initial visual target parameters of the visual target based on the LSTM model specifically includes: inputting the historical data of the historical tester into the LSTM model, where the historical data includes the behavioral data, the corresponding visual target information, and the test results of the historical tester set in time series; outputting the prediction parameters of the visual target according to the historical data; and adjusting the initial visual target parameters based on the prediction parameters, where the initial visual target parameters include the size and direction of the visual target.

[0012] Further, if the tester is a historical tester, before dynamically adjusting the visual target based on the LSTM model, the method further includes: constructing the LSTM model, where the LSTM model includes a model input layer, an LSTM layer, and a fully connected layer; inputting the historical data of the historical tester into the model input layer; processing the historical data through the LSTM layer to obtain the dynamic changes of the historical tester; mapping the dynamic changes output by the LSTM layer to the prediction parameters of the visual target through the fully connected layer; and optimizing the prediction parameters by using a regression loss function.

[0013] In addition, the present invention also provides an intelligent vision testing system, including: a display device; at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the intelligent vision testing method described above.

[0014] Furthermore, the present invention also provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the intelligent vision testing method described above.

[0015] Compared with the prior art, the intelligent vision detection method provided by the present invention creatively calculates the target visual angle and size of the visual target dynamically according to the test distance; generates the visual target by using the equivalent vision method; and displays the visual target on the display device with the target visual angle being consistent. The present invention integrates the dynamic visual target generation technology of the equivalent visual angle method into vision detection, which is an important technical means to achieve the consistency of the equivalent visual angle during the vision detection process. This technology can be unrestricted by distance when conforming to GB / T 11533-2011 according to the actual detection distance between the detection device and the tester; meanwhile, this method can realize single-person self-service vision testing, dynamically calculate and adjust the size and shape of the visual target in real time to ensure that the visual angle formed by the visual target is exactly the same as the detection conditions of standard vision, and the detection accuracy is high. The intelligent vision detection method is not restricted by space and can be applied to vision detection scenarios in small spaces. The intelligent vision testing method can file the behavior data, corresponding visual target information, and test results of the tester, and automatically analyze the changes in the historical data of the tester, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings constituting a part of the present invention creatively are used to provide a further understanding of the present invention creatively. The schematic embodiments of the present invention creatively and their descriptions are used to explain the present invention creatively and do not constitute an improper limitation to the present invention creatively. In the drawings: Figure 1Schematic flowchart of an intelligent vision detection method provided by an embodiment of the present invention; Figure 2 For Figure 1 Schematic diagram of displaying an eye chart on a display device in the intelligent vision detection method shown; Figure 3 Schematic flowchart of a specific intelligent vision detection method provided by another embodiment of the present invention; Figure 4 Schematic diagram of an intelligent vision detection system provided by an embodiment of the present invention. Detailed implementation manners

[0017] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. Similar elements in different embodiments are labeled with related similar element numbers. In the following embodiments, many detailed descriptions are provided to enable a better understanding of the present invention. However, those skilled in the art can easily recognize that some of the features can be omitted in different situations, or can be replaced by other elements, materials, or methods. In some cases, some operations related to the present invention are not shown or described in the specification, which is to avoid the core part of the present invention being overwhelmed by excessive descriptions. For those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations based on the descriptions in the specification and general technical knowledge in the art.

[0018] It should be noted that, without conflict, the embodiments and features in the embodiments of the present invention can be combined with each other to form various embodiments. At the same time, the steps or actions in the method description can also be reordered or adjusted in an obvious manner by those skilled in the art. Therefore, the various sequences in the specification and drawings are only for clearly describing a certain embodiment, and do not mean that they are necessary sequences, unless it is stated that a certain sequence must be followed.

[0019] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention. In addition, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first", "second", etc. may explicitly or implicitly include one or more of such features. In the description of the present invention, unless otherwise stated, the meaning of "a plurality" is two or more.

[0020] In the description of the present invention, it should be noted that unless otherwise clearly defined and limited, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood through specific situations.

[0021] The present invention will be described in detail below with reference to the drawings and in conjunction with embodiments.

[0022] As Figure 1 shown, an embodiment of the present invention provides an intelligent vision detection method, including the following steps: S10: Dynamically calculate the target visual angle and size of the visual target according to the test distance; S20: Generate the visual target by using the equivalent vision method; S30: Display the visual target on the display device while maintaining the consistency of the target visual angle.

[0023] The embodiment of the present invention integrates the dynamic visual target generation technology of the equivalent visual angle method into vision detection, which is an important technical means to achieve the consistency of the equivalent visual angle in the vision detection process. This technology can dynamically calculate and adjust the size and shape of the visual target in real time according to the actual detection distance between the device and the tester, ensuring that the visual angle formed by the visual target is exactly the same as that under the standard vision detection conditions, and improving the accuracy of the detection result. This intelligent vision detection method can be applied to small spaces, and the tester can independently complete vision detection without being restricted by the space distance.

[0024] When the intelligent vision testing method provided by the embodiments of the present invention complies with GB / T 11533-2011, it is not restricted by the spatial distance. At the same time, it can realize single-person self-service vision testing. The present invention integrates the equivalent viewing angle method dynamic visual target generation technology into vision detection, which is an important technical means to achieve the consistency of equivalent viewing angles during the vision detection process.

[0025] This intelligent vision detection method can ensure that the visual target has a consistent visual angle (viewing angle) at different test distances by dynamically calculating the size of the visual target using the equivalent vision method and making real-time adjustments according to the test distance.

[0026] Further, in the intelligent vision detection method provided by the embodiments of the present invention, S10: Dynamically calculate the target viewing angle and size of the visual target according to the test distance, specifically including: Dynamically calculate the viewing angle of the visual target according to the viewing angle formula; Select the optimal one from the viewing angles of the visual target as the target viewing angle; Determine the size of the visual target based on the target viewing angle.

[0027] The core goal of generating and displaying the visual target is to ensure that the visual target has a consistent viewing angle (i.e., visual angle) at different test distances during the vision testing process. By dynamically calculating the size of the visual target and making real-time adjustments according to the test distance, the optimal visual target can be obtained. The viewing angle of the optimal visual target obtained will be used as the target viewing angle.

[0028] Further, in the intelligent vision detection method provided by the embodiments of the present invention, S20: Generate the visual target using the equivalent vision method, specifically including: Calculate the equivalent viewing angle of the visual target based on the size of the visual target: (1) Among them, is the equivalent viewing angle (in radians), is the pixel size of the visual target on the display device (in pixels), is the test distance (in centimeters), is the pixel density of the display device (in pixels per inch), and 2.54 is the conversion factor for converting inches to centimeters. Among them, the pixel size of the visual target on the display device : (2) The pixel density PPI of the display device: (3) Among them, is the actual physical size of the visual target (in centimeters), is the resolution width of the display device, is the resolution height of the display device, is the diagonal size of the display device (in inches).

[0029] Usually, the distance between the tester and the testing device varies. According to the equivalent viewing angle method, the viewing angle formula listed in formula (4) is used to calculate the viewing angle of the visual target. And the size of the visual target is calculated dynamically in real time according to formula (5). During this testing process, the size of the visual target needs to be adjusted dynamically according to the testing distance, and it is ensured that the viewing angle of the visual target always remains consistent.

[0030] Dynamically calculate and generate the visual target according to the viewing angle formula (4) to ensure that each generated visual target remains consistent.

[0031] (4) Among them, is the actual size of the visual target (usually the width or height of the visual target), is the testing distance between the visual target and the observer.

[0032] Then, determine the target viewing angle according to formula (5). In order to ensure that each generated visual target has the same viewing angle, a target viewing angle needs to be determined first . Then, we reverse-derive and calculate the size of the visual target through formula (5) so that the target viewing angle is , and this step needs to be completed when calculating the various sizes and opening directions of the visual target.

[0033] (5) Finally, generate the visual target. According to the calculated the actual size of the visual target can be obtained, and then it is applied to the front-end display device. In actual operation, the generated visual target can be a single letter, symbol, and the sizes of these patterns, considering the pixels of the display device, the formulas of the equivalent viewing angle method are as in formulas (1), (2), (3). Through the above steps, the function of generating and displaying the visual target can provide consistent and accurate visual acuity measurement results in different tests, ensuring the accuracy and reliability of the visual acuity test.

[0034] By adopting the equivalent vision method to generate the visual target, the visual target on the display device can be dynamically adjusted at different testing distances to ensure that the visual target is displayed in the correct size on the display device. Through the above formulas, the physical size of the visual target can be converted into the pixel size on the display device to ensure that the visual target is displayed in the correct size on the display device.

[0035] The intelligent vision test method provided by the embodiments of the present invention can make the visual target adapt to different test distances, and can dynamically adjust the size of the visual target by using the equivalent visual angle technology. When testing at a short distance, the actual size of the visual target is larger and the visual angle is larger because the distance between the observer and the display device is smaller. When testing at a long distance, the actual size of the visual target is smaller and the visual angle is smaller because the distance between the observer and the display device is larger. In practical applications, the active light-emitting display device can quickly and accurately adjust the display size of the visual target. The visual target can be adjusted in real time according to different test distances without manual intervention by the tester. The size of the visual target is dynamically calculated based on the current test distance and the PPI of the display device, and then the display content of the visual target is adjusted to ensure that the visual target can maintain a consistent visual angle under different test conditions.

[0036] Further, in the intelligent vision detection method provided by the embodiments of the present invention, after step S30: displaying the visual target on the display device with the target visual angle being consistent, it further includes: receiving and storing the behavior data and the corresponding visual target information made by the tester, so as to output the prediction parameters of the visual target based on the behavior data and the visual target information during the next test. It should be noted that both the initial visual target parameters and the prediction parameters mentioned in this application include the direction and size of the visual target. The storage of relevant tester information of the tester can save the relevant data of the tester in the vision test for each tester, including personal information, behavior data, corresponding visual target information, test results, etc. It can help the tester track their own vision changes, analyze the vision trend, and provide customized tests and suggestions when needed. The data of each test will be stored in the database in chronological order. The specific data storage format may include: Tester ID (unique identifier) Test time Tester's vision value At the beginning of each new test, it is necessary to first query the stored historical data to judge the vision change trend or other performances during the test of the tester. These historical data will help the LSTM (Long Short-Term Memory Network) dynamically adjust the parameters of the visual target. The data storage also supports the update function: after each new test is completed, the latest test result of the tester will be appended to the historical data to ensure the real-time nature of the data.

[0037] The visual target generated for the first time when the tester undergoes the first vision detection should be the visual target corresponding to 4.0 in the 5-point record of GB / T 11533-2011. After the tester makes the initial selection, record the correctness of the selection. Dynamically generate the visual target in real time and actually record the number of times the visual target appears during the inspection process. After measuring 10-15 times in total, stop the test, and calculate the vision value according to the tester's selection during the test. The vision value calculation method adopts the vision averaging method of GB / T 11533-2011.

[0038] Further, in the intelligent vision detection method provided by the embodiment of the present invention, in S30: displaying the visual target on the display device while keeping the target perspective consistent, specifically including: if the cumulative number of incorrect judgments on the visual target direction by the tester reaches the set threshold, then display the size of the visual target that was correctly judged last time, and reset the number of incorrect judgments.

[0039] Further, in the intelligent vision detection method provided by the embodiment of the present invention, before receiving and storing the behavior data and the corresponding visual target information made by the tester on the visual target, it further includes: if the tester is a historical tester, dynamically adjusting the initial visual target parameters of the visual target based on the Long Short-Term Memory (LSTM) model.

[0040] The intelligent vision testing method provided by the embodiment of the present invention applies the Long Short-Term Memory (LSTM) to the dynamic adjustment of the visual target, and can intelligently predict the visual target according to the behavior data of the historical tester set in the historical time series, and adjust the parameters of the visual target such as size and direction, so as to provide a personalized vision testing experience for each tester. LSTM can process time series data and adapt to the reaction mode of the tester in real time, optimize the testing process, and improve the accuracy and effect of vision examination.

[0041] Further, in the intelligent vision detection method provided by the embodiment of the present invention, if the tester is a historical tester, dynamically adjusting the initial visual target parameters of the visual target based on the LSTM model, specifically including: inputting the historical data of the historical tester into the LSTM model, where the historical data includes the behavior data, the corresponding visual target information, and the test results of the historical tester set in the time series; outputting the predicted parameters of the visual target according to the historical data; and adjusting the initial visual target parameters based on the predicted parameters, where the initial visual target parameters include the size and direction of the visual target.

[0042] Using the LSTM model to analyze the tester's data in the vision test, the LSTM algorithm can dynamically adjust the size and direction parameters of the visual target based on the historical data of the tester during the vision test. LSTM can analyze the reaction mode of the tester during the test (for example, whether the tester often repeats choosing incorrect answers, whether the reaction time is long, etc.), and dynamically adjust the visual target based on these reactions. And the vision performance and adaptability of each tester are different, and LSTM can predict and generate a personalized visual target test sequence by analyzing the historical data of the tester.

[0043] Visual acuity tests are not always independent. When conducting a visual acuity examination, the tester's response is not only related to the current visual target but may also be related to their past response patterns. By analyzing the tester's past behavioral data, such as factors like the trend of historical measurements, LSTM can discover the implicit patterns in these time-series historical data and adjust the parameters of the current visual target based on these patterns.

[0044] First, design the input data for the LSTM model. The LSTM model is a deep learning model suitable for processing time-series data. Therefore, in a visual acuity test, the input data needs to include the tester's historical data set in a time series, which can capture the dynamic changes of the tester during the test. The characteristics of the input historical data can include: Behavioral data: 1. The tester's response behavior; 2. The tester's response time: the time it takes for the tester to select a visual target. For example, a longer time to select a visual target may indicate that the tester has poorer eyesight or needs more time to adapt. The historical data also includes the test results: 1. The correctness of visual target selection: the record of the tester's correct and incorrect selections in the visual acuity test. An incorrect selection may mean poorer eyesight or that the tester was not fully focused on the visual acuity test; 2. The test difficulty: the difficulty of the visual target, which can also include parameters such as the clarity and contrast of the visual target; 3. The visual acuity value for each visual acuity test.

[0045] The historical data also includes the corresponding visual target information: 1. The visual target size: the size of the visual target in the current visual acuity test, which can be letters, numbers, or other forms of visual targets. 2. The direction of the visual target: the direction of the visual target may affect the tester's response, and the change in the direction of the visual target needs to be recorded.

[0046] The historical data also includes the tester's historical visual acuity data, the results of the tester's previous visual acuity tests, including whether there are eye diseases, etc.

[0047] Then, design the output data of the LSTM model. The LSTM model predicts the predicted parameters (direction and size) of the visual target in the current or future visual acuity test based on the input historical data, and adjusts the visual target according to the predicted parameters during the visual acuity test. The predicted parameters are the predicted visual target size and direction suitable for the tester's current test difficulty output by the LSTM model based on the tester's historical data. For example, some testers perform poorly on rotating visual targets, and normal-oriented visual targets can be output for these testers.

[0048] Furthermore, in the intelligent vision detection method provided by the embodiments of the present invention, if the tester is a historical tester, before dynamically adjusting the visual target based on the LSTM model, the method further includes: constructing an LSTM model, which includes a model input layer, an LSTM layer, and a fully connected layer; inputting the historical data of the historical tester into the model input layer; processing the historical data through the LSTM layer to obtain the dynamic changes of the historical tester; mapping the dynamic changes output by the LSTM layer to the prediction parameters of the visual target through the fully connected layer; and optimizing the prediction parameters using a regression loss function.

[0049] When constructing the LSTM model and training the model, data set collection and preprocessing can be performed. Collect a large amount of vision test data of testers, including the historical data of testers. Then, collect these vision test data, label the behavioral data of testers (correct or incorrect selections, response time, response speed, etc.) and the visual target information, and use them as the target input for model training. Normalization: Standardize the input data to ensure the balance of data during model training. In particular, normalize numerical features such as the response time and selection correctness of testers. Model input layer: Input historical data such as the historical response data of testers (such as time, error rate, etc.) and the relevant parameters of the visual target into the model input layer. The LSTM layer processes the historical data set with time series settings to obtain the dynamic changes of the historical tester. The fully connected layer maps the dynamic changes output by the LSTM layer to the prediction parameters, such as the visual target size, direction, etc. Use a regression loss function to optimize the prediction parameters of the visual target, such as the predicted value of the visual target size. The tester conducts a test at a certain time. For this test at this time point, n different visual targets will be generated for the tester to judge and select to test the user's vision value. And the directions and vision values of these n visual targets are generated by the LSTM model. The tester can conduct multiple tests at multiple time periods to detect the change of the vision value. In this case, the tester is called a historical tester, and the test results at each time period are historical data. If the test results of the tester at multiple time periods are different, it means that the vision value of this tester has changed. The LSTM can intelligently analyze the changes in historical data, and the changes in this historical data are dynamic changes.

[0050] Use historical data to train the LSTM model, iteratively optimize the parameters of the LSTM model, and gradually adjust the prediction parameters of the visual target output by the model. During training, the gradient descent method or the Adam optimizer can be used, and cross-validation can be used to avoid overfitting.

[0051] By building an LSTM model and applying it to the dynamic adjustment of visual acuity test optotypes, a personalized visual acuity test experience is provided for each tester. By processing historical data set according to time series, the optotypes in the visual acuity test process are dynamically adjusted to adapt to the tester's response mode in real time, optimizing the visual acuity test process, and ultimately improving the accuracy and effectiveness of visual acuity examination. Specific embodiments See Figure 3 The following is a specific flowchart of an intelligent visual acuity test system provided by an embodiment of the present invention.

[0053] After the system starts, the primary task is to identify the tester ID and query information in the database in depth. If the tester does not exist, the system quickly reports an error and terminates the process to ensure the standardization of the test process and the accuracy of the data. If the tester exists, it is further determined whether the tester is a historical tester. For historical testers, the system extracts their previous test results and predicts the initial parameters of the optotypes for this test with the help of the LSTM model; for non-historical testers, the initial visual acuity value is directly set to 4.0 and the direction is random.

[0054] Counter A in the intelligent visual acuity test system mainly controls the total number of tests, and it is stipulated that the number of measurements is 30 times. When counter A reaches 30, the test is basically completed and directly enters the visual acuity value calculation link to avoid overtesting. Counter B is used to control the number of attempts when the tester selects the wrong optotype direction. When the tester selects the wrong option and counter B ≤ 2, the tester is allowed to continue to try, and at the same time counter A and B increase; if counter B > 2, it means that the tester has tried too many times under the current size optotype. The system generates the previous size optotype and resets counter B = 1, and predicts the next optotype parameters of this test process by combining the tester's selection result sequence of this test process with the LSTM model, and gives the tester a more appropriate difficulty optotype to continue the test, ensuring the rationality of the test and the test experience.

[0055] In the optotype display stage, after the user selects the optotype direction, the system makes decisions on the follow-up based on the correctness of the selection and the values of counters A and B. When the selection is correct and counters A and B meet the conditions, the optotype of the current size is generated; when the selection is wrong, it is processed according to the status of counter B. After the number of tests reaches the standard, the system calculates the visual acuity value by the average method of visual acuity in GB / T 11533 - 2011, stores the test information, and ends the process.

[0056] Embodiment 2 Correspondingly, according to an embodiment of the present invention, an intelligent visual acuity test system, a readable storage medium, and a computer program product are also provided. Figure 4An intelligent vision test system provided in an embodiment of the present invention includes a display device and a data processor. In terms of software and data processing, this intelligent vision test system can generate visual targets that comply with the national standard GB / T 11533-2011 and implement functions such as display, analysis of the behavioral data of the tester, calculation of vision values, and storage of relevant information of the tester. The display device can be an actively emitting display device. When presenting visual targets, it is necessary to ensure the clear display of the visual targets, that is, even in the case of extremely small fonts or graphics, all details can be kept recognizable. Especially when simulating different vision levels, it is necessary to ensure that even the visual targets with the smallest size and the slightest direction difference can be clearly presented. A high-resolution display device can provide richer display details, ensuring that visual targets of various sizes and directions can be clearly presented without blurring or distortion due to resolution limitations. The data processor has powerful computing capabilities and graphics rendering performance, and can quickly analyze the data input by the tester. Through the precise processing of these data, the best visual target pattern can be generated, which is further used as the visual target for the tester's next vision examination, thereby providing a more accurate vision health assessment.

[0057] It should be understood that various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in the disclosure of the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and no limitations are imposed herein.

[0058] The above specific embodiments do not constitute a limitation to the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent vision detection method, characterized in that: It includes the following steps: Dynamically calculate the target viewing angle and size of the visual acuity chart according to the test distance; Generate the visual acuity chart by using the equivalent vision method; Display the visual acuity chart on the display device with consistency in the target viewing angle.

2. The intelligent vision detection method according to claim 1, wherein: Dynamically calculate the target viewing angle and size of the visual acuity chart according to the test distance, specifically including: Dynamically calculate the viewing angle of the visual acuity chart according to the viewing angle formula; Select the optimal one from the viewing angles of the visual acuity chart as the target viewing angle; Determine the size of the visual acuity chart based on the target viewing angle.

3. The intelligent vision detection method according to claim 2, characterized in that: Generate the visual acuity chart by using the equivalent vision method, specifically including: Calculate the equivalent viewing angle of the visual acuity chart based on the size of the visual acuity chart: （1） Among them, is the equivalent viewing angle (unit: radian), is the pixel size of the visual target on the display device (unit: pixel), is the test distance (unit: centimeter), is the pixel density of the display device (unit: pixel / inch), and 2.54 is the conversion factor for converting inches to centimeters. Among them, the pixel size of the visual target on the display device : （2） The pixel density PPI of the display device: （3） wherein, is the actual physical size of the visual target (unit: centimeter), is the resolution width of the display device, is the resolution height of the display device, is the diagonal size of the display device (unit: inch).

4. The intelligent vision detection method according to any one of claims 1 to 3, characterized in that: After displaying the visual acuity chart on the display device with consistency in the target viewing angle, the method further includes: Receive and store the behavioral data and corresponding visual acuity chart information made by the tester, so as to output the prediction parameters of the visual acuity chart based on the behavioral data and the visual acuity chart information during the next test.

5. The intelligent vision detection method according to claim 4, wherein: Display the visual acuity chart on the display device with consistency in the target viewing angle, specifically including: If the cumulative number of incorrect judgments on the viewing direction of the visual acuity chart by the tester reaches the set threshold, then display the size of the visual acuity chart that was correctly judged last time, and reset the number of incorrect judgments.

6. The intelligent vision detection method according to claim 4, characterized in that: Before receiving and storing the behavioral data and corresponding visual acuity chart information made by the tester, the method further includes: If the tester is a historical tester, dynamically adjust the initial visual acuity chart parameters based on the long short-term memory network (LSTM) model.

7. The intelligent vision detection method according to claim 6, wherein: If the tester is a historical tester, dynamically adjust the initial visual acuity chart parameters based on the LSTM model, specifically including: Input the historical data of the historical tester into the LSTM model, where the historical data includes the behavioral data, corresponding visual acuity chart information, and test results of the historical tester set in time series; Output the prediction parameters of the visual acuity chart according to the historical data; Adjust the initial visual acuity chart parameters based on the prediction parameters, where the initial visual acuity chart parameters include the size and direction of the visual acuity chart.

8. The intelligent vision detection method according to claim 6, characterized in that: Before dynamically adjusting the visual acuity chart based on the LSTM model if the tester is a historical tester, the method further includes: Construct the LSTM model, which includes a model input layer, an LSTM layer, and a fully connected layer; Input the historical data of the historical tester into the model input layer; Process the historical data through the LSTM layer to obtain the dynamic changes of the historical tester; Map the dynamic changes output by the LSTM layer to the prediction parameters of the visual acuity chart through the fully connected layer; Optimize the prediction parameters by using a regression loss function.

9. An intelligent vision testing system, characterized in that: It includes: A display device; At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the intelligent visual acuity test method according to any one of claims 1 to 8.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the intelligent vision test method according to any one of claims 1 to 8.

Images