Analysis method and system based on strabismus type

Abstract

The invention discloses an analysis method and system based on strabismus types, and belongs to the technical field of intelligent medical treatment. Comprising the following steps: acquiring video data through an acquisition terminal, and transmitting the video data to a processing terminal; the processing terminal receives and processes the video data to obtain processed detection data; the detection data comprises short-distance detection data and long-distance detection data; and analyzing the detection data by adopting an image processing algorithm, and judging whether a strabismus event occurs or not and judging the strabismus type by calculating the change of central points of pupils of the left and right eyes in an alternate covering state. Compared with the prior art, the method has the advantages that errors and inconsistency possibly caused by subjective judgment depending on experience of professionals in a traditional method are avoided, and even slight or difficult-to-perceive strabismus events can be recognized through intelligent analysis.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical technology, and more specifically, to an analysis method and system based on strabismus types. Background Art

[0002] As a special eye condition, the diversity of strabismus is reflected in various types such as latent strabismus, esotropia, exotropia, as well as more complex vertical strabismus and cyclotropia. These different types of strabismus not only affect the individual's appearance perception, but more importantly, they may have a profound impact on the normal function and development of the visual system, and may thus lead to problems such as a decline in visual acuity level and the absence of stereoscopic vision ability.

[0003] In the field of strabismus detection and analysis, the existing technologies have shown obvious limitations. Traditional detection methods, such as synoptophore examination, prism examination, and cover-uncover test, although widely used in practical applications, highly rely on the experience and judgment of professionals, and the operation process is relatively complex and time-consuming. These detection methods pose relatively high requirements for the cooperation of the tested person, requiring the tested person to be able to accurately understand and execute instructions, which poses a significant challenge for children, especially preschool children, who are young and have difficulty concentrating their attention for a long time.

[0004] Specifically, the existing strabismus detection technologies have the following main defects: (1) High equipment cost: Traditional detection methods usually require the use of expensive professional equipment, such as synoptophore and prism, which not only increases the overall cost of detection, but also limits the popularization scope of detection services to a certain extent; (2) Strong subjectivity: The accuracy and consistency of detection results largely depend on the personal experience and subjective judgment of the detection personnel, which may lead to differences in results among different detection personnel; (3) High cooperation requirement: For the child group, especially young children, due to their limited cognitive and self-control abilities, it is difficult for them to maintain concentrated attention for a long time and cooperate to complete complex detection processes, which may lead to inaccurate detection results or data loss; (4) Long detection cycle: Traditional detection methods often require multiple detections and comprehensive evaluations by combining different test means, which not only extends the time cycle from detection to result output, but also increases the waiting cost of the tested person.

[0005] As can be seen from the above, the related technologies do not give any technical inspiration on how to perform more efficient, accurate, and adaptable strabismus type analysis. Summary of the Invention

[0006] 1. Technical Problem to be Solved

[0007] In view of the problem in the prior art of how to analyze the strabismus type more efficiently, accurately and adaptively, the present invention provides an analysis method and system based on the strabismus type, which can combine image processing technology and intelligent algorithms to deeply analyze eye images, so as to accurately identify the strabismus type.

[0008] 2. Technical solution

[0009] The object of the present invention is achieved by the following technical solutions.

[0010] The content part of this application is used to introduce the concept in a brief form, and these concepts will be described in detail in the subsequent specific implementation part. The content part of this application is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0011] Some embodiments of this application propose an analysis method and system based on the strabismus type to solve the technical problems mentioned in the above background art part.

[0012] As the first aspect of this application, some embodiments of this application provide an analysis method based on the strabismus type, including the following steps: obtaining video data through an acquisition terminal and transmitting it to a processing terminal; the processing terminal receiving and processing the video data to obtain processed detection data; the detection data includes close-range detection data and long-range detection data; using an image processing algorithm to analyze the detection data, and by calculating the changes in the center points of the left and right eye pupils under the alternate occlusion state, determining whether a strabismus event occurs and the strabismus type.

[0013] Furthermore, the acquisition terminal includes the first micro camera, the second micro camera and a liquid crystal light valve on the left and right sides;

[0014] When the acquisition terminal shoots the binocular movement trajectory video, the shooting duration each time is t, and the shooting interval frequency is w;

[0015] Furthermore, the process of alternate occlusion is realized by switching the power-on and power-off states of the liquid crystal light valve.

[0016] Furthermore, the process of obtaining video data includes close-range detection and long-range detection, and the video data includes the binocular movement trajectory video of close-range detection and the binocular movement trajectory video of long-range detection;

[0017] The processing of the video data includes the processing of the binocular movement trajectory video obtained by close-range detection and the processing of the binocular movement trajectory video obtained by long-range detection.

[0018] Further, during the near-distance detection, a black-and-white square target is used as the near-distance target, and the distance between the subject and the near-distance target is at least 33 cm.

[0019] Further, the near-distance detection process includes: turning off both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve, keeping both eyes fixed on the near-distance target for a duration of h1;

[0020] Switching the left-eye liquid crystal light valve to the powered-on state to cover the left eye, keeping the right eye fixed on the near-distance target for a duration of h2;

[0021] Switching both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve to the powered-off state, removing the cover, and keeping both eyes fixed on the near-distance target for a duration of h3;

[0022] Switching the right-eye liquid crystal light valve to the powered-on state to cover the right eye, keeping the left eye fixed on the near-distance target for a duration of h2;

[0023] Switching both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve to the powered-off state, removing the cover, and keeping both eyes fixed on the near-distance target for a duration of h3.

[0024] Further, the process of obtaining near-distance detection data includes: according to the binocular movement trajectory video obtained from the near-distance detection, when both eyes are fixed on the near-distance target, obtaining the coordinates of the center point of the left pupil and the coordinates of the center point of the right pupil;

[0025] When the left eye is covered, obtaining the coordinates of the center point of the right pupil and the coordinates of the center point of the left pupil;

[0026] After removing the cover of the left eye, obtaining the coordinates of the center point of the left pupil and the coordinates of the center point of the right pupil;

[0027] When the right eye is covered, obtaining the coordinates of the center point of the left pupil and the coordinates of the center point of the right pupil;

[0028] After removing the cover of the right eye, obtaining the coordinates of the center point of the left pupil and the coordinates of the center point of the right pupil.

[0029] Further, during the far-distance detection, an LED lamp group is used as the far-distance target, and the distance between the subject and the far-distance target is L1.

[0030] Further, the process of analyzing the detection data includes: setting a threshold value, monitoring the detection data frame by frame to trigger the determination of a strabismus event;

[0031] Combining the near-distance detection data and the far-distance detection data, identifying the strabismus type through a machine learning model;

[0032] Data smoothing and interpolation processing are introduced to exclude abnormal interference, and the binocular coordinate differences in the alternating occlusion state are calculated to determine the strabismus type.

[0033] As a second aspect of the present application, some embodiments of the present application provide an analysis system based on strabismus type, which is characterized by including an acquisition terminal and a processing terminal;

[0034] The acquisition terminal acquires video data and sends it to the processing terminal; the processing terminal receives the video data and processes it to obtain detection data and analyze it to determine whether a strabismus event occurs and the strabismus type.

[0035] 3. Beneficial effects

[0036] Compared with the prior art, the advantages of the present invention are as follows:

[0037] (1) By ingeniously integrating a high-precision infrared camera and a liquid crystal light valve in the acquisition terminal, the system of the present invention can instantaneously capture and record the fine eye movement trajectories of both eyes in the near and far states. Compared with the traditional methods that rely on large and fixed subjective detection devices (such as visual acuity charts, prism testing devices, etc.), the present invention not only greatly reduces the space limitation, but also eliminates the cumbersome operation steps. The person to be detected only needs to simply wear or approach the acquisition device to complete the detection, which greatly improves the convenience and instantaneity of the detection and is applicable to more scenarios;

[0038] (2) The eye movement trajectory data collected by the present invention is directly stored in a portable or cloud processing terminal, which allows users to access their eye health records at any time, and can review and analyze historical data to achieve continuous tracking and trend analysis of the eye health status. This is crucial for early detection of potential problems such as vision decline and strabismus development, helps to take timely intervention measures, and effectively delays or reverses vision problems;

[0039] (3) The present invention uses image processing algorithms to deeply analyze the recorded video data, can automatically identify the subtle features of eye movements, and can accurately distinguish different types of strabismus events (such as esotropia, exotropia, hypertropia or hypotropia). This automated process improves the accuracy and efficiency of judging strabismus events, avoids the errors and inconsistencies that may be brought by relying on the experience of professionals for subjective judgment in traditional methods, and can identify even minor or imperceptible strabismus events through intelligent analysis, which is particularly important for high-risk groups of strabismus such as children. Brief description of the drawings

[0040] Figure 1 It is a flowchart of an analysis method based on strabismus type in an embodiment of the present invention;

[0041] Figure 2Flow chart of near - distance detection and far - distance detection in an embodiment of the present invention;

[0042] Figure 3 Overall structure diagram of the analysis system based on strabismus type in an embodiment of the present invention.

[0043] Explanation of reference numerals in the figure: 1. Infrared lamp; 2. Miniature camera; 3. Nose pad fixing hole; 4. Eyeglass lens; 5. Inner frame of glasses; 6. Main frame of glasses; 7. Temple; 8. Data connection line; 9. Processing terminal; 10. Liquid crystal light valve. Detailed implementation manners

[0044] The present invention will be described in detail below with reference to the accompanying drawings of the specification and specific embodiments.

[0045] As Figure 1 shown, an analysis system based on strabismus type of the present invention includes a collection terminal and a processing terminal 9. The collection terminal is used to obtain video data and send it to the processing terminal 9; the processing terminal 9 is used to receive and process the video data, obtain detection data and perform analysis to determine whether a strabismus event occurs and the type of strabismus, and finally output the analysis result.

[0046] In a specific embodiment, the collection terminal can be a glasses device, and the processing terminal 9 can be a cloud processing terminal 9 or a local processing terminal 9. The glasses device includes an infrared lamp 1, a miniature camera 2, a liquid crystal light valve 10, an eyeglass lens 4, an inner frame of glasses 5, a main frame of glasses 6, and a temple 7.

[0047] Specifically, the miniature camera 2 is arranged at the bottom of the main frame 6 of the glasses. The miniature camera 2 includes a first miniature camera 2 and a second miniature camera 2. The first miniature camera 2 is arranged at the center of the bottom of the left frame of the main frame 6 of the glasses, and the second miniature camera 2 is arranged at the center of the bottom of the right frame of the main frame 6 of the glasses. And an infrared filter is arranged on one side of each camera. The infrared filter plays a role in protecting the miniature camera 2 and filtering infrared rays. Through the setting of the infrared filter, it can selectively absorb or reflect infrared light of a specific wavelength to filter out infrared rays and only allow visible light to pass through, thereby improving the imaging quality.

[0048] Specifically, the included angle between the miniature camera 2 and the direction perpendicular to the lens is accurately set to 34° to 36°, and the infrared filter also maintains this angle range. This ensures that the camera can accurately capture the images of the user's both eyes, including fine details such as the corners of the eyes and the upper and lower eyelids.

[0049] In a specific embodiment, the included angle between the micro camera 2 and the direction perpendicular to the lens is 35°. At the same time, the included angle between the infrared filter and the direction perpendicular to the lens is 35°. When the included angle is set to 35°, the system can obtain high-definition binocular images of child and adolescent users, meeting the detection requirements for specific age groups. The spectacle lens 4 is used for refractive compensation and can be made of various materials such as glass lenses, resin lenses, PC lenses or crystal lenses to meet the personalized needs of different users.

[0050] The distance between the inner frame 5 of the glasses and the pupil is set to 7 mm to 15 mm, reserving space for installing refractive compensation lenses to correct the myopia, hyperopia or astigmatism problems of users.

[0051] The main frame 6 of the glasses is made of TR (plastic titanium) material. A plurality of infrared lamps 1 are evenly arranged on the main frame 6 of the glasses, providing a lighting source for the micro camera 2 to monitor the pupil without affecting the scaling of the pupil.

[0052] Specifically, the infrared lamps 1 are arranged at the bottom of the main frame 6 of the glasses, enabling clear binocular images to be captured even in low-light environments. The main frame 6 of the glasses and the temple 7 can be equipped with customized anti-slip sleeves, which can be adjusted slightly back and forth according to the wearer's head shape to ensure that the main body of the glasses is stable and does not shake, improving the safety and comfort of wearing.

[0053] More specifically, the processing terminal 9 can be a processing terminal 9 device using an Intel Core i7 processor. The processing terminal 9 includes a user management module, a calibration module, a storage module and an analysis module. The user management module is used to enable users to perform personalized settings on the terminal and the acquisition terminal, including managing user information, controlling the switch of the camera, setting the recording cycle and frequency, etc.; the calibration module is used to correct the wearing position of the acquisition terminal by analyzing the eye image data collected by the acquisition terminal, which ensures that the detected person has correctly worn the acquisition terminal and provides an accurate data basis for subsequent strabismus analysis; the storage module is used to store the binocular motion images recorded by the acquisition terminal in real time to ensure the integrity and traceability of the data; the analysis module uses image processing algorithms to deeply analyze the stored image data to determine whether there is strabismus and further identify the specific type of strabismus. At the same time, the analysis module can also monitor the working status of the camera in real time to ensure the accuracy and reliability of the data.

[0054] In a specific embodiment, the acquisition terminal and the processing terminal 9 can be electrically connected through a Bluetooth module or a data cable. After the acquisition terminal enters the working state, the processing terminal 9 will immediately receive and process the eye image data collected by the acquisition terminal through the micro camera 2. If the acquisition terminal is worn incorrectly or the camera cannot work properly, the processing terminal 9 will promptly send a prompt message to guide the user to make adjustments.

[0055] Specifically, the processing terminal 9, through the calibration module, corrects the position of the acquisition terminal according to the acquired eye image data collected by the acquisition terminal, so as to ensure that the detected person has correctly worn the eye end. During the process of the detected person wearing the eye end, the processing terminal 9 analyzes whether the eye end has been correctly worn through the analysis module and issues a prompt message, thereby prompting the detected person to adjust the eye end to achieve correct wearing.

[0056] Correctly wearing the eye end means that the micro camera 2 of the eye end can completely acquire the images of the corners of the eyes, the upper and lower eyelids, and the pupils.

[0057] The analysis system based on the strabismus type of the present invention realizes the accurate detection and analysis of the strabismus phenomenon through the close cooperation between the acquisition terminal and the processing terminal 9. The system not only has high convenience and accuracy, but also provides personalized setting options and intelligent analysis functions, providing users with all-round eye health monitoring services.

[0058] This technical solution integrates a high-precision infrared camera and a liquid crystal light valve in the acquisition terminal, realizing the instant capture and recording of the fine eye movement trajectories of both eyes at close and far distances. It not only greatly reduces the space limitation, simplifies the operation steps, improves the convenience and instantaneity of detection, and is applicable to more scenarios; at the same time, the collected data is directly stored in a portable or cloud processing terminal, facilitating users to access the eye health files at any time, realizing continuous tracking and trend analysis, and helping to detect vision problems early and intervene in a timely manner.

[0059] Combined Figures 1 to 3 , an analysis method based on the strabismus type of the present invention includes the following steps:

[0060] S1. Video data acquisition:

[0061] Acquire the binocular movement trajectory video through a wearable acquisition terminal, and transmit the acquired video data to the processing terminal for subsequent analysis.

[0062] Specifically, the binocular movement trajectory video is captured by the micro cameras of the acquisition terminal. The micro cameras of the acquisition terminal include a first micro camera and a second micro camera located on the left and right sides of the main frame of the glasses. The duration of each capture of the binocular movement trajectory video is t, and the frequency is w, where t ≥ 1 minute and w ≥ 1 second / time, to ensure the continuity and integrity of the data, and at the same time effectively manage the data volume and processing time.

[0063] In a specific embodiment, the duration t of each capture of the binocular movement trajectory video is 1 minute, and the frequency w is 1 capture every 1 second.

[0064] The process of video data acquisition includes near-distance detection and far-distance detection. The video data includes the binocular movement trajectory videos of near-distance detection and far-distance detection.

[0065] More specifically, during the shooting process of the binocular movement trajectory videos, near-distance alternating occlusion shooting and far-distance alternating occlusion shooting are adopted to complete near-distance detection and far-distance detection, so as to obtain the binocular movement trajectory videos of near-distance detection and far-distance detection, which are used to comprehensively evaluate the eye alignment state and potential strabismus problems.

[0066] During near-distance alternating occlusion shooting, the near distance is set to L1. The subject gazes at a square target with a side length of 11.53 mm at near distance, so as to simulate near-distance visual activity scenarios such as daily reading and writing. During far-distance alternating occlusion shooting, the far distance is set to L2. The subject gazes at a target object containing multiple light sources (such as the light emitted by a flashlight) at far distance, so as to simulate the far-distance scenario of daily visual activities. Then, the binocular movement trajectory videos of near-distance detection and far-distance detection captured are sent to the processing terminal to complete storage and detection.

[0067] In a specific embodiment, during near-distance alternating occlusion shooting, the near distance L1 is 133 cm, and during far-distance alternating occlusion shooting, the far distance L2 is 6 m.

[0068] The process of alternating occluder shooting is realized by the power-on and power-off of the liquid crystal light valve: the process of alternating occlusion shooting is realized by the state switching of the power-on and power-off of the liquid crystal light valve. When the liquid crystal light valve is in the power-on state, it switches to the fully black opaque mode, effectively blocking the light from passing through. When the liquid crystal light valve is in the power-off state, it switches to the transparent mode, and the light transmittance exceeds 80%, allowing normal vision.

[0069] The near-distance detection process includes: setting a standardized near-distance target, making both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve in the power-off state, keeping both eyes gazing at the near-distance target and lasting for a duration of h1, where h1≥3 seconds;

[0070] Switch the left-eye liquid crystal light valve to the power-on state to achieve left-eye occlusion, keep the right eye gazing at the near-distance target and last for a duration of h2, where h2≥10 seconds;

[0071] Switch both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve to the power-off state to remove the occlusion, keep both eyes gazing at the near-distance target and last for a duration of h3, where h3≥5 seconds;

[0072] Switch the right-eye liquid crystal light valve to the power-on state to achieve right-eye occlusion, keep the left eye gazing at the near-distance target and last for a duration of h2;

[0073] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state, remove the occlusion, and keep both eyes fixating on the near - distance visual target for a duration of h3.

[0074] In a specific embodiment, the near - distance detection process is as follows:

[0075] Set a standardized near - distance visual target, and fixedly place a black - and - white square visual target with a side length of 11.53 mm at a distance of 33 cm from the person being tested.

[0076] Keep both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve in the power - off state. At this time, it is in the transparent mode. Keep both eyes fixating on the near - distance visual target at a distance of 33 cm for 3 seconds.

[0077] Switch the left - eye liquid - crystal light valve to the power - on state to achieve occlusion of the left eye, and keep the right eye continuing to fixate on the near - distance visual target for 10 seconds.

[0078] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state, remove the occlusion, and keep both eyes fixating on the near - distance visual target for 5 seconds.

[0079] Switch the right - eye liquid - crystal light valve to the power - on state to achieve occlusion of the right eye, and keep the left eye continuing to fixate on the near - distance visual target for 10 seconds.

[0080] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state, remove the occlusion, and keep both eyes fixating on the near - distance visual target for 5 seconds.

[0081] The far - distance detection process includes: Set a standardized far - distance visual target, keep both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve in the power - off state, keep both eyes fixating on the near - distance visual target for a duration of h1, where h1≥3 seconds;

[0082] Switch the left - eye liquid - crystal light valve to the power - on state to achieve occlusion of the left eye, keep the right eye fixating on the near - distance visual target for a duration of h2, where h2≥10 seconds;

[0083] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state, remove the occlusion, and keep both eyes fixating on the near - distance visual target for a duration of h3, where h3≥5 seconds;

[0084] Switch the right - eye liquid - crystal light valve to the power - on state to achieve occlusion of the right eye, keep the left eye fixating on the near - distance visual target for a duration of h2;

[0085] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state, remove the occlusion, and keep both eyes fixating on the near - distance visual target for a duration of h3.

[0086] In a specific embodiment, the far - distance detection process is as follows:

[0087] Set a standardized long - distance visual target, and arrange multiple LED light groups at a distance of 6 meters from the person to be detected (total illuminance ≥ 100 lux, color temperature 5000K ± 200K);

[0088] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state. At this time, it is in the transparent mode. Keep both eyes fixating on the visual target at a distance of 6 meters and continue for 3 seconds;

[0089] Switch the left - eye liquid - crystal light valve to the power - on state, and keep the right - eye liquid - crystal light valve in the power - off state to achieve left - eye occlusion. Keep the right eye continuing to fixate on the long - distance visual target for 10 seconds;

[0090] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state, remove the occlusion, and keep both eyes fixating on the long - distance visual target for 5 seconds;

[0091] Switch the right - eye liquid - crystal light valve to the power - on state, and keep the left - eye liquid - crystal light valve in the power - off state to achieve right - eye occlusion. Keep the left eye continuing to fixate on the long - distance visual target for 10 seconds;

[0092] Switch both the left - eye liquid - crystal light valve and the right - eye liquid - crystal light valve to the power - off state, remove the occlusion, and keep both eyes fixating on the long - distance visual target for 5 seconds.

[0093] Thus, through near - distance detection and long - distance detection, obtain the binocular movement trajectory videos during the near - distance detection process and the binocular movement trajectory videos during the long - distance detection process, and send them as video data to the processing terminal for data processing and analysis.

[0094] S2. Detection data processing:

[0095] After the processing terminal receives the video data collected by the acquisition terminal, it processes the video data to obtain the processed detection data, and then judges whether there is strabismus and the specific type of strabismus.

[0096] Specifically, the processing process of the video data includes the processing of the binocular movement trajectory video obtained during the near - distance detection process and the processing of the binocular movement trajectory video obtained during the long - distance detection process. The detection data includes near - distance detection data and long - distance detection data, both of which focus on obtaining the pupil center point coordinates in the binocular movement trajectory video.

[0097] In a specific embodiment, the processing process of the binocular movement trajectory video during the near - distance detection is as follows:

[0098] According to the binocular movement trajectory video obtained during the near - distance detection process, when both eyes are fixating on the near - distance visual target, obtain the pupil center point coordinate (CBLZhu) when the left eye fixates on the near - distance visual target and the pupil center point coordinate (CBRZhu) when the right eye fixates on the near - distance visual target;

[0099] When the left eye is covered, obtain the pupil center point coordinates (CRZhu) when the right eye fixates on a near target and the pupil center point coordinates (CLZhe) when the left eye is covered at this time;

[0100] After removing the left eye cover, obtain the left eye pupil center point coordinates (CLL) and the right eye pupil center point coordinates (CLR);

[0101] Similarly, when the right eye is covered, obtain the pupil center point coordinates (CLzhu) when the left eye fixates on a near target and the pupil center point coordinates (CRzhe) when the right eye is covered;

[0102] After removing the right eye cover, obtain the left eye pupil center point coordinates (CRL) and the right eye pupil center point coordinates (CRR).

[0103] Thus, obtain the near - distance detection data.

[0104] In a specific embodiment, the processing procedure for the binocular movement trajectory video during the long - distance detection is as follows:

[0105] According to the binocular movement trajectory video obtained during the long - distance detection, when both eyes fixate on a long - distance target simultaneously, obtain the pupil center point coordinates (FBLZhu) when the left eye fixates on a near target and the pupil center point coordinates (FBRZhu) when the right eye fixates on a near target;

[0106] When the left eye is covered, obtain the pupil center point coordinates (FRZhu) when the right eye fixates on a long - distance target and the pupil center point coordinates (FLZhe) when the left eye is covered at this time;

[0107] After removing the left eye cover, obtain the left eye pupil center point coordinates (FLL) and the right eye pupil center point coordinates (FLR);

[0108] Similarly, when the right eye is covered, obtain the pupil center point coordinates (FLzhu) when the left eye fixates on a long - distance target and the pupil center point coordinates (FRzhe) when the right eye is covered;

[0109] After removing the right eye cover, obtain the left eye pupil center point coordinates (FRL) and the right eye pupil center point coordinates (FRR).

[0110] Thus, obtain the long - distance detection data.

[0111] After completing the processing of the binocular movement trajectory videos during the near-distance detection and the binocular movement trajectory videos during the far-distance detection, near-distance detection data and far-distance detection data are obtained. The system will conduct a detailed analysis of each coordinate value in the acquired detection data, especially comparing the coordinate differences of the pupil center points between the covered eye and the uncovered eye. By comprehensively evaluating these differences and combining medical standards and algorithm models, the system can accurately determine whether a strabismus event has occurred and further identify the type of strabismus. This process not only considers the alignment of the two eyes at different distances but also deeply analyzes the changes in pupil positions during the alternating cover test, thus ensuring the accuracy and reliability of strabismus detection.

[0112] S3. Detection data analysis:

[0113] After obtaining the detection data, image processing algorithms are used to analyze the processed detection data. The core lies in calculating the changes in the pupil center points of the left and right eyes during the alternating cover test, comprehensively judging whether a strabismus event has occurred, and further determining the specific type of strabismus. The specific steps for detecting data analysis are as follows:

[0114] (1) Threshold setting and preliminary judgment:

[0115] Set the offset thresholds, including the pupil center point offset threshold and the speed threshold. The pupil center point offset threshold is the maximum allowable offset of the pupil center point in different states, and the speed threshold is the maximum value of the offset speed.

[0116] Specifically, monitor the detection data frame by frame or continuously for multiple frames. Once the change in the pupil center point in a certain frame or consecutive frames of the binocular movement trajectory video data exceeds the preset threshold range, the system will immediately determine that a strabismus event has occurred.

[0117] (2) Type identification and secondary judgment:

[0118] Combining the near-distance detection data and the far-distance detection data, comprehensively evaluate the overall change trend of the pupil center points of the left and right eyes. At the same time, analyze the relative position relationship between the pupil center points of the left and right eyes in different states. To more accurately identify the type of strabismus, a machine learning model is introduced to identify various types of strabismus, including constant exotropia, intermittent exotropia, latent strabismus, and esotropia.

[0119] During the identification process, especially to consider excluding misjudgments that may be caused by abnormal situations such as glasses displacement, detachment, blinking, and closing eyes, this embodiment adopts data smoothing and interpolation processing techniques to enhance the robustness of the image processing algorithm to abnormal data, thus ensuring the accuracy and reliability of the analysis results.

[0120] Specifically, after identifying and eliminating interference factors through an image processing algorithm, the differences between other values and the left and right eye fixation eyes during near-distance detection and far-distance detection are calculated respectively. Combining these differences, it is comprehensively determined whether a strabismus event occurs, and the type of strabismus is further analyzed. Finally, based on detailed data analysis and statistics, an accurate judgment result of the strabismus type is obtained.

[0121] In a specific embodiment, as shown in Table 1, the judgment conditions for analyzing strabismus events and types based on near-distance detection and far-distance detection are used to clearly define the occurrence of strabismus events and the types of strabismus.

[0122] Table 1 Judgment Conditions for Strabismus Events and Types

[0123]

[0124] Specifically, based on the judgment conditions shown in Table 1 and according to the near-distance detection data and far-distance detection data, the process of judging constant exotropia specifically includes:

[0125] S301. Judge whether the difference between the left eye CBLzhu before unocclusion and the left dominant eye CLZhu after covering the right eye is less than -5 prism diopters;

[0126] If so, continue with the judgment in S302;

[0127] If not, end the judgment;

[0128] S302. Judge whether the difference between the right eye CBRzhu before unocclusion and the right dominant eye CRZhu after covering the left eye is less than -5 prism diopters;

[0129] If so, continue with the judgment in S303;

[0130] If not, end the judgment;

[0131] S303. Judge whether the differences between the binocular eyes when removing the right eye occlusion and the binocular fixation eyes before occlusion, namely CLL - CLZhu and CLR - CRZhu, are both less than -5 prism diopters;

[0132] If so, continue with the judgment in S304;

[0133] If not, end the judgment;

[0134] S304. Judge whether the differences between the binocular eyes when removing the left eye occlusion and the binocular fixation eyes before occlusion, namely CRL - CLZhu and CRR - CRZhu, are both less than -5 prism diopters;

[0135] If so, continue with the judgment in S305;

[0136] If not, end the judgment.

[0137] S305. Determine whether the difference between the left eye FBLzhu before occlusion and the left dominant eye FLZhu after covering the right eye is less than -5 prism diopters;

[0138] If so, continue with the judgment in S306;

[0139] If not, end the judgment;

[0140] S306. Determine whether the difference between the right eye FBRzhu before occlusion and the right dominant eye FRZhu after covering the left eye is less than -5 prism diopters;

[0141] If so, continue with the judgment in S307;

[0142] If not, end the judgment;

[0143] S307. Determine whether the differences between the binoculars after removing the right eye occlusion and the binocular fixation eyes before occlusion, namely FLL - FLZhu and FLR - FRZhu, are both less than -5 prism diopters;

[0144] If so, continue with the judgment in S308;

[0145] If not, end the judgment;

[0146] S308. Determine whether the differences between the binoculars after removing the left eye occlusion and the binocular fixation eyes before occlusion, namely FRL - FLZhu and FRR - FRZhu, are both less than -5 prism diopters;

[0147] If so, the judgment result is constant exotropia;

[0148] If not, end the judgment.

[0149] In a specific embodiment, based on the judgment conditions shown in Table 1, according to the near - distance detection data and the far - distance detection data, the process of judging intermittent exotropia specifically includes:

[0150] S311. Determine whether the difference between the left eye CBLzhu before occlusion and the left dominant eye CLZhu after covering the right eye is less than -5 prism diopters;

[0151] If so, judge as intermittent exotropia and end the judgment;

[0152] If not, continue with the judgment in S312;

[0153] S312. Determine whether the difference between the right eye CBRzhu before occlusion and the right dominant eye CRZhu after covering the left eye is less than -5 prism diopters;

[0154] If so, it is determined as intermittent exotropia and the determination ends;

[0155] If not, continue with the determination in S313;

[0156] S313. Determine whether the difference between covering the left eye CLZhe and single fixation on the left eye CLZhu after covering the right eye is less than -10 prism diopters;

[0157] If so, it is determined as intermittent exotropia and the determination ends;

[0158] If not, continue with the determination in S314;

[0159] S314. Determine whether the difference between single fixation on the right eye after covering the left eye and covering the right eye is less than -10 prism diopters;

[0160] If so, it is determined as intermittent exotropia and the determination ends;

[0161] If not, continue with the determination in S315;

[0162] S315. Determine whether the differences between binoculars after removing the right eye cover and the binocular fixation eye before covering, namely CLL - CLZhu and CLR - CRZhu, are both less than -5 prism diopters;

[0163] If so, it is determined as intermittent exotropia and the determination ends;

[0164] If not, continue with the determination in S316;

[0165] S316. Determine whether the differences between binoculars after removing the left eye cover and the binocular fixation eye before covering, namely CRL - CLZhu and CRR - CRZhu, are both less than -5 prism diopters;

[0166] If so, it is determined as intermittent exotropia and the determination ends;

[0167] If not, continue with the determination in S317;

[0168] S317. Determine whether the difference between the left eye FBLzhu before covering and the left dominant eye FLZhu after covering the right eye is less than -5 prism diopters;

[0169] If so, it is determined as intermittent exotropia and the determination ends;

[0170] If not, continue with the determination in S318;

[0171] S318. Determine whether the difference between the right eye FBRzhu before covering and the right dominant eye FRZhu after covering the left eye is less than -5 prism diopters;

[0172] If so, it is determined as intermittent exotropia and the determination ends;

[0173] If not, continue with the determination in S319;

[0174] S319. Determine whether the difference between covering the left eye FLZhe and the left main fixation eye FLZhu after covering the right eye is less than -10 prism diopters;

[0175] If so, it is determined as intermittent exotropia and the determination ends;

[0176] If not, continue with the determination in S320;

[0177] S320. Determine whether the difference between the right main fixation eye after covering the left eye and covering the right eye is less than -10 prism diopters;

[0178] If so, it is determined as intermittent exotropia and the determination ends;

[0179] If not, continue with the determination in S321;

[0180] S321. Determine whether the differences between the binocular eyes after removing the right eye cover and the binocular fixation eyes before covering, namely FLL - FLZhu and FLR - FRZhu, are both less than -5 prism diopters;

[0181] If so, it is determined as intermittent exotropia and the determination ends;

[0182] If not, continue with the determination in S322;

[0183] S322. Determine whether the differences between the binocular eyes after removing the left eye cover and the binocular fixation eyes before covering, namely FRL - FLZhu and FRR - FRZhu, are both less than -5 prism diopters;

[0184] If so, it is determined as intermittent exotropia and the determination ends;

[0185] If not, end the determination.

[0186] In a specific embodiment, based on the judgment conditions shown in Table 1, the process of judging normal or esophoria according to the near - distance detection data and the far - distance detection data specifically includes:

[0187] S331. Determine whether the difference between the left eye CBLzhu before covering and the left main fixation eye CLZhu after covering the right eye is greater than -5 prism diopters;

[0188] If so, continue with the determination in S332;

[0189] If not, end the determination;

[0190] S332. Determine whether the difference between the right eye CBRzhu before occlusion and the right dominant eye CRZhu after covering the left eye is greater than -5 prism diopters;

[0191] If so, continue with the judgment in S333;

[0192] If not, end the judgment;

[0193] S333. Determine whether the difference between the left eye CLZhe and the single fixation left eye CLZhu after covering the right eye is greater than -10 prism diopters;

[0194] If so, continue with the judgment in S334;

[0195] If not, end the judgment;

[0196] S334. Whether the difference between the single fixation right eye after covering the left eye and the covered right eye is greater than -10 prism diopters. If so, continue with the judgment in S335; if not, end the judgment;

[0197] S335. Whether the differences between the binoculars with the right eye uncovered and the binocular fixation eyes before occlusion, that is, CLL - CLZhu and CLR - CRZhu, are both greater than -5 prism diopters;

[0198] If so, continue with the judgment in S336;

[0199] If not, end the judgment;

[0200] S336. Whether the differences between the binoculars with the left eye uncovered and the binocular fixation eyes before occlusion, that is, CRL - CLZhu and CRR - CRZhu, are both greater than -5 prism diopters;

[0201] If so, continue with the judgment in S337;

[0202] If not, end the judgment.

[0203] S337. Determine whether the difference between the left eye FBLzhu before occlusion and the left dominant eye FLZhu after covering the right eye is greater than -5 prism diopters;

[0204] If so, continue with the judgment in S338;

[0205] If not, end the judgment;

[0206] S338. Determine whether the difference between the right eye FBRzhu before occlusion and the right dominant eye FRZhu after covering the left eye is greater than -5 prism diopters;

[0207] If so, continue with the judgment in S339;

[0208] If not, end the judgment;

[0209] S339. Determine whether the difference between covering the left eye FLZhe and the single fixation on the left eye FLZhu after covering the right eye is greater than -10 prism diopters;

[0210] If so, continue with the judgment in S340;

[0211] If not, end the judgment;

[0212] S340. Determine whether the difference between the single fixation on the right eye after covering the left eye and covering the right eye is greater than -10 prism diopters;

[0213] If so, continue with the judgment in S341;

[0214] If not, end the judgment;

[0215] S341. Determine whether the differences between the binocular eyes after removing the right eye cover and the binocular fixation eyes before covering, namely FLL - FLZhu and FLR - FRZhu, are both greater than -5 prism diopters;

[0216] If so, continue with the judgment in S342;

[0217] If not, end the judgment;

[0218] S342. Determine whether the differences between the binocular eyes after removing the left eye cover and the binocular fixation eyes before covering, namely FRL - FLZhu and FRR - FRZhu, are both greater than -5 prism diopters;

[0219] If so, the judgment result is normal or exophoria, and end the judgment;

[0220] If not, end the judgment.

[0221] In a specific embodiment, based on the judgment conditions shown in Table 1, the process of judging normal or esophoria according to the near - distance detection data and the far - distance detection data specifically includes:

[0222] S351. Determine whether the difference between the left eye CBLzhu before covering and the left dominant eye CLZhu after covering the right eye is less than 5 prism diopters;

[0223] If so, continue with the judgment in S352;

[0224] If not, end the judgment;

[0225] S352. Determine whether the difference between the right eye CBRzhu before covering and the right dominant eye CRZhu after covering the left eye is less than 5 prism diopters;

[0226] If so, continue with the judgment in S353;

[0227] If not, end the judgment;

[0228] S353. Determine whether the differences between the binoculars with the right eye covered and the binocular fixation eyes before covering, that is, CLL - CLZhu and CLR - CRZhu, are both less than 5 prism diopters;

[0229] If so, continue with the judgment in S354;

[0230] If not, end the judgment;

[0231] S354. Determine whether the differences between the binoculars with the left eye covered and the binocular fixation eyes before covering, that is, CRL - CLZhu and CRR - CRZhu, are both less than 5 prism diopters;

[0232] If so, continue with the judgment in S355;

[0233] If not, end the judgment;

[0234] S355. Determine whether the difference between the left eye FBLzhu before covering and the left dominant eye FLZhu after covering the right eye is less than 5 prism diopters;

[0235] If so, continue with the judgment in S356;

[0236] If not, end the judgment;

[0237] S356. Determine whether the difference between the right eye FBRzhu before covering and the right dominant eye FRZhu after covering the left eye is less than 5 prism diopters;

[0238] If so, continue with the judgment in S357;

[0239] If not, end the judgment;

[0240] S357. Determine whether the differences between the binoculars with the right eye covered and the binocular fixation eyes before covering, that is, FLL - FLZhu and FLR - FRZhu, are both less than 5 prism diopters;

[0241] If so, continue with the judgment in S358;

[0242] If not, end the judgment;

[0243] S358. Determine whether the differences between the binoculars with the left eye covered and the binocular fixation eyes before covering, that is, FRL - FLZhu and FRR - FRZhu, are both less than 5 prism diopters;

[0244] If so, the judgment result is normal or esophoria, and end the judgment;

[0245] If not, end the judgment.

[0246] In a specific embodiment, based on the judgment conditions shown in Table 1, the process of judging esotropia according to the near-distance detection data and the far-distance detection data specifically includes:

[0247] S361. Judge whether the difference between the left eye CBLzhu before occlusion and the left dominant eye CLZhu after covering the right eye is greater than 5 prism diopters;

[0248] If so, the judgment result is esotropia, and end the judgment;

[0249] If not, continue to make the judgment of S362;

[0250] S362. Judge whether the difference between the right eye CBRzhu before occlusion and the right dominant eye CRZhu after covering the left eye is greater than 5 prism diopters;

[0251] If so, the judgment result is esotropia, and end the judgment;

[0252] If not, continue to make the judgment of S363;

[0253] S363. Judge that the comparison between the left eye CLZhe when covering the left eye and the single fixation on the left eye CLZhu after covering the right eye is greater than 10 prism diopters;

[0254] If so, the judgment result is esotropia, and end the judgment;

[0255] If not, continue to make the judgment of S364;

[0256] S364. Judge whether the difference between the single fixation on the right eye after covering the left eye and the covered right eye is greater than 10 prism diopters;

[0257] If so, the judgment result is esotropia, and end the judgment;

[0258] If not, continue to make the judgment of 365;

[0259] S365. Judge whether the differences between the binoculars when removing the right eye occlusion and the binocular fixation eyes before occlusion, that is, CLL-CLZhu and CLR-CRZhu, are both greater than 5 prism diopters;

[0260] If so, the judgment result is esotropia, and end the judgment;

[0261] If not, continue to make the judgment of 366;

[0262] S366. Judge whether the differences between the binoculars when removing the left eye occlusion and the binocular fixation eyes before occlusion, that is, CRL-CLZhu and CRR-CRZhu, are both greater than 5 prism diopters;

[0263] If so, the judgment result is esotropia, and the judgment ends;

[0264] If not, continue with the judgment in 367;

[0265] If not, the judgment result is non - esotropia; if so, the judgment result is esotropia, and the judgment ends.

[0266] S367. Is the difference between the left eye FBLzhu before uncovering and the left dominant eye FLZhu after covering the right eye greater than 5 prism diopters?

[0267] If so, the judgment result is esotropia, and the judgment ends;

[0268] If not, continue with the judgment in 368;

[0269] S368. Is the difference between the right eye FBRzhu before uncovering and the right dominant eye FRZhu after covering the left eye greater than 5 prism diopters?

[0270] If so, the judgment result is esotropia, and the judgment ends;

[0271] If not, continue with the judgment in 369;

[0272] S369. Are the differences between the binocular fixation eyes with the right eye uncovered and before uncovering, namely FLL - FLZhu and FLR - FRZhu, both greater than 5 prism diopters?

[0273] If so, the judgment result is esotropia, and the judgment ends;

[0274] If not, continue with the judgment in 370;

[0275] S370. Are the differences between the binocular fixation eyes with the left eye uncovered and before uncovering, namely FRL - FLZhu and FRR - FRZhu, both greater than 5 prism diopters?

[0276] If so, the judgment result is esotropia, and the judgment ends;

[0277] If not, the judgment ends.

[0278] The present invention integrates a micro - camera array with inclination and an intelligent liquid - crystal light valve dynamic covering system in a wearable acquisition terminal to realize the automatic execution of the covering test. Compared with the traditional manual detection, the efficiency is greatly improved, especially solving the problem of difficult detection caused by the distracted attention of the children's group.

[0279] Secondly, by establishing a pixel coordinate - prism diopter conversion algorithm and a multi - dimensional judgment matrix, the precise quantitative analysis of the pupil displacement and movement trajectory in the cover test is realized. Compared with the traditional manual interpretation, the accuracy of identifying intermittent exotropia in this solution has been improved by...

[0280] Furthermore, by using image - processing algorithms to deeply analyze video data, the subtle features of eye movement are automatically identified and the types of strabismus are accurately distinguished, improving the accuracy and efficiency of judging strabismus events and avoiding the errors that may be brought by traditional subjective judgments, which is particularly important for high - risk groups of strabismus.

[0281] In addition, the cloud - based collaborative processing architecture reduces the equipment cost. Combined with the abnormal data compensation mechanism, it realizes the universal detection without the need for a professional site, providing continuous data support for the establishment of individualized eye health records. The above schematically describes the present invention and its implementation manners. This description is not restrictive. Without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. What is shown in the drawings is only one of the implementation manners of the present invention, and the actual structure is not limited thereto. Any reference signs in the claims should not limit the claimed rights. Therefore, if those of ordinary skill in the art are inspired by it and, without departing from the purpose of this creation, design similar structural ways and embodiments to this technical solution without creative efforts, they should all fall within the protection scope of this patent. In addition, the word "comprising" does not exclude other elements or steps, and the word "a" before an element does not exclude including "a plurality" of such elements. The multiple elements stated in the product claims can also be implemented by one element through software or hardware. Words such as first, second, etc. are used to indicate names and do not indicate any specific order.

Claims

1. An analysis method based on strabismus type, comprising the following steps: Obtain video data through a collection terminal and transmit it to a processing terminal; The processing terminal receives and processes the video data to obtain processed detection data; the detection data includes close-range detection data and long-range detection data; Use an image processing algorithm to analyze the detection data, and judge whether a strabismus event occurs and the strabismus type by calculating the changes in the center points of the left and right eye pupils under the alternating occlusion state.

2. The analysis method based on strabismus type according to claim 1, wherein: The collection terminal includes a first micro camera, a second micro camera and a liquid crystal light valve on the left and right sides; When the collection terminal shoots a video of the binocular movement trajectory, the shooting duration each time is t, and the frequency of the shooting interval is w.

3. The analysis method based on strabismus type according to claim 1, wherein: The process of alternating occlusion is realized by switching the power-on and power-off states of the liquid crystal light valve.

4. The analysis method based on strabismus type according to claim 1, wherein: The process of obtaining video data includes close-range detection and long-range detection, and the video data includes a binocular movement trajectory video for close-range detection and a binocular movement trajectory video for long-range detection; The processing of the video data includes the processing of the binocular movement trajectory video obtained by close-range detection and the processing of the binocular movement trajectory video obtained by long-range detection.

5. The analysis method based on strabismus type according to claim 4, wherein: During the close-range detection process, a black and white square target is used as the close-range target, and the distance between the person to be detected and the close-range target is at least 33 cm.

6. The analysis method based on strabismus type according to claim 4, wherein: The close-range detection process includes: making both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve in the power-off state, keeping both eyes fixed on the close-range target and lasting for a duration of h1; Switch the left-eye liquid crystal light valve to the power-on state to achieve left-eye occlusion, keep the right eye fixed on the close-range target and last for a duration of h2; Switch both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve to the power-off state to remove the occlusion, keep both eyes fixed on the close-range target and last for a duration of h3; Switch the right-eye liquid crystal light valve to the power-on state to achieve right-eye occlusion, keep the left eye fixed on the close-range target and last for a duration of h2; Switch both the left-eye liquid crystal light valve and the right-eye liquid crystal light valve to the power-off state to remove the occlusion, keep both eyes fixed on the close-range target and last for a duration of h3.

7. The analysis method based on strabismus type according to claim 4, wherein: The process of obtaining close-range detection data includes: according to the binocular movement trajectory video obtained by close-range detection, when both eyes are fixed on the close-range target, obtain the coordinates of the center point of the left eye pupil and the coordinates of the center point of the right eye pupil; When the left eye is occluded, obtain the coordinates of the center point of the right eye pupil and the coordinates of the center point of the left eye pupil; After removing the left-eye occlusion, obtain the coordinates of the center point of the left eye pupil and the coordinates of the center point of the right eye pupil; When the right eye is occluded, obtain the coordinates of the center point of the left eye pupil and the coordinates of the center point of the right eye pupil; After removing the right eye occlusion, obtain the coordinates of the center point of the left pupil and the coordinates of the center point of the right pupil.

8. The analysis method based on the type of strabismus according to claim 4, characterized in that: During the long-distance detection process, use the LED lamp group as the long-distance visual target, and the distance between the person to be detected and the long-distance visual target is L1.

9. The analysis method based on the type of strabismus according to claim 1, characterized in that: The process of analyzing the detection data includes: setting a threshold, and monitoring the detection data frame by frame to trigger the determination of the strabismus event; Combining the near-distance detection data and the long-distance detection data, and identifying the type of strabismus through a machine learning model; Introduce data smoothing and interpolation processing to exclude abnormal interference, calculate the binocular coordinate difference in the alternating occlusion state, and judge the type of strabismus.

10. An analysis system based on the type of strabismus, characterized in that, It includes a collection terminal and a processing terminal; The collection terminal obtains video data and sends it to the processing terminal; The processing terminal receives the video data and processes it to obtain detection data and analyze it to determine whether a strabismus event occurs and the type of strabismus.

Images