Intermittent external strabismus dynamic monitoring system and method based on upper and lower frame judgment

Abstract

The invention relates to the technical field of eye position detection, in particular to an intermittent external strabismus dynamic monitoring system and method based on upper and lower frame judgment, and the system comprises an uploading module which is used for uploading a user eye movement image and decomposing the user eye movement image into user eye movement image frames; the selection module is used for traversing the user eye movement picture frames obtained by decomposition in the uploading module, and selecting not less than three user eye movement picture frames on each section of the eye movement path; according to the method, the user eye movement picture frame is captured in the user eye movement image, and the planned eye movement path and the actual eye movement path of the user are acquired by taking the user eye movement picture frame as a basic parameter, so that whether the eye position of the user has the outward inclination defect or not is judged on the basis of similarity comparison of the two types of eye movement paths; the eye movement monitoring technology is applied to the maximum extent to assist medical staff to diagnose the outward inclination problem of the eyes of the user more quickly and accurately.

Description

Technical Field

[0001] The present invention relates to the technical field of eye position detection, and particularly relates to an intermittent exotropia dynamic monitoring system and method based on upper and lower frame judgment. Background Art

[0002] Eye position exotropia is a common ophthalmic disease, manifested as the eyes deviating outwards. It may be caused by abnormal development of extraocular muscles, nerve problems, regulatory factors, etc. Mild exotropia may only affect appearance. When the degree of eye position exotropia is large, it will cause diplopia, amblyopia, and also affect stereoscopic vision, and can be treated by surgical or non-surgical methods.

[0003] The invention patent application with the application number 201910532640.X discloses a learning evaluation method based on eye movement recognition, including: establishing a learning evaluation model; the learning evaluation model includes an eye movement recognition information setting value and a corresponding evaluation value to obtain eye movement recognition information; matching the eye movement recognition information with the eye movement recognition information setting value of the learning evaluation model to obtain a corresponding evaluation value; the eye movement recognition information is obtained in the following manner: obtaining at least a pair of eyeball images, the eyeball images including a sclera region and a pupil region: performing binarization processing on the eyeball images: the gray value of the pupil region is greater than the gray value of the sclera region, and determining the moving direction of the pupil region according to the pixel change of the sclera region; obtaining the corresponding evaluation value includes: when obtaining the switching time point of the teaching content screen and the time t after switching, judging frame by frame whether each pair of eyeball images has the corresponding eyeball movement action of the teaching content screen; matching the obtained eyeball movement information with the established learning evaluation model to obtain a corresponding evaluation value.

[0004] This application aims to solve the problem: "Currently, in the field of human-computer interaction technology, there have been many attempts at emerging interaction methods, such as somatosensory interaction, eye movement tracking, voice interaction, biometric identification, etc. However, the usage rates of most interaction methods are not very high, and they have not yet entered the real commercial application popularization. Moreover, there is no human-computer interaction method that can achieve the level where people can communicate with devices without any obstacles and at will. Eye movement research is the main means to explore the laws of human attention and cognition: it records the real movement of the eyeballs, describes people's visual behavior, and reflects people's cognitive processing and psychological activities".

[0005] However, when the eye movement monitoring technology is used in the field of eye position exotropia disease diagnosis, although it can relatively accurately monitor the eye movement of patients, it cannot further diagnose whether there is an exotropia defect in the user's eye position based on the monitored data.

[0006] Therefore, an intermittent exotropia dynamic monitoring system and method based on upper and lower frame judgment are proposed. Summary of the Invention

[0007] In view of the above-mentioned drawbacks of the prior art, the present invention provides an intermittent exotropia dynamic monitoring system and method based on upper and lower frame judgment, which solves the technical problems proposed in the above-mentioned background art.

[0008] To achieve the above objectives, the present invention is realized through the following technical solutions:

[0009] In a first aspect, an intermittent exotropia dynamic monitoring system based on upper and lower frame judgment includes:

[0010] An upload module for uploading user eye movement images and decomposing the user eye movement images into user eye movement frame pictures; a selection module for traversing the user eye movement frame pictures decomposed in the upload module and selecting no less than three user eye movement frame pictures on each path of the eye movement path; an identification module for receiving the user eye movement frame pictures selected by the operation of the selection module, picking up reference points in the user eye movement frame pictures, and identifying the coordinates of the reference points in each user eye movement frame picture; a construction module for receiving the reference point coordinates identified by the operation of the identification module and constructing a real eye movement path based on the reference point coordinates; a comparison module for comparing the similarity between each corresponding real eye movement path and the eye movement path.

[0011] During the operation stage of the comparison module, identify the marked time sequence of the user eye movement frame picture from which the end point of the real eye movement path comes, determine a time domain based on the two marked time sequences, and intercept a local eye movement path of the same time domain in the corresponding eye movement path of the real eye movement path with the determined time domain. Further:

[0012] ;

[0013] Where: is the comprehensive similarity between each real eye movement path and the eye movement path; is the total amount of real eye movement paths; is the length and slope of the i-th real eye movement path; is the length and slope of the eye movement path corresponding to the i-th real eye movement path; is an adjustment factor;

[0014] Among them, the adjustment factor takes a value of 1 or -1. The numerator of the fraction where the adjustment factor is located is less than or equal to the denominator. The adjustment factor takes a value of 1. Otherwise, the adjustment factor takes a value of -1;

[0015] A determination module, configured to set an external strabismus determination threshold for eye position, obtain the similarity obtained by comparison in the comparison module, and when the similarity is not less than the external strabismus determination threshold for eye position, determine that the user has an external strabismus defect in the eye position; otherwise, determine that the user does not have an external strabismus defect in the eye position:

[0016] The uploading module is wirelessly interactively connected to a selection module. The lower level of the selection module is wirelessly interactively connected to an optimization unit and a sub-packaging unit. The selection module is wirelessly interactively connected to an identification module. The identification module is wirelessly interactively connected to the sub-packaging unit. The identification module is wirelessly interactively connected to a construction module, a comparison module, and a determination module.

[0017] Furthermore, the user eye movement image uploaded in the uploading module is sourced from an eye movement image acquisition device, which is integrated by a fixed camera, a marker, and an electric slide rail. The marker is installed on the electric slide rail, and the movement of the marker is controlled by the electric slide rail. During the user eye movement image acquisition stage, the user is prompted to always fixate on the marker, and the marker moves along a predetermined path by the electric slide rail. The fixed camera completes the acquisition of the user eye movement image during the process of the electric slide rail controlling the movement of the marker and uploads it to the uploading module for storage;

[0018] Among them, there are no less than three bending nodes in the predetermined path used by the electric slide rail to control the movement of the marker, and the frame rate of the fixed camera for acquiring the user eye movement image is not less than 60 Hz.

[0019] Furthermore, each user eye movement frame obtained by decomposing in the uploading module is marked with its time sequence in the user eye movement image. The eye movement path is the predetermined path used by the electric slide rail to control the movement of the marker. The user eye movement image is segmented based on the timestamps of the start and end of each segment of the path in the eye movement path, obtaining several sub-user eye movement images. Each sub-user eye movement image corresponds to a set of user eye movement frames, and each set of user eye movement frames is used as the target for the selection module to select user eye movement frames, and the selection of user eye movement frames is performed;

[0020] After the user eye movement frames selected by the selection module are sorted based on the marked time sequence, the number of intervening frames between each adjacent user eye movement frame is equal, and the first frame and the last frame in the sorted user eye movement frames respectively correspond to the start timestamp and the end timestamp of the source sub-user eye movement image of the frame;

[0021] Among them, the higher the requirement for the diagnostic accuracy of the external strabismus of the user's eye position, the more bending nodes in the predetermined path and the more user eye movement frames selected by the selection module; otherwise, the fewer.

[0022] Furthermore, a sub-module is provided at the lower level of the selection module, including:

[0023] Optimization unit, configured to receive the user eye movement video frames selected by the selection module and perform optimization processing on the user eye movement video frames;

[0024] Packet splitting unit, configured to receive the user eye movement video frames that have completed optimization processing in the optimization unit, read the marker time sequence of the user eye movement video frames, determine the source sub-user eye movement image of the user eye movement video frames, and based on the source sub-user eye movement image of the user eye movement video frames, distinguish the user eye movement video frames to obtain a plurality of user eye movement video frame data packets;

[0025] Among them, an optimization processing logic for user eye movement video frames is set in the optimization unit. The optimization unit performs optimization processing on the user eye movement video frames based on the optimization processing logic for user eye movement video frames. The user eye movement video frames included in the user eye movement video frame data packets are sorted and placed based on their respective marker time sequences.

[0026] Furthermore, the optimization processing logic of the user eye movement video frames is expressed as:

[0027] ;

[0028] In the formula: is the pixel value of the finally optimized user eye movement video frame at (x, y); represents the histogram equalization operation; represents the guided filter operation; is the gamma coefficient;

[0029] Among them, in the stage of the guided filter operation, both inputs are the enhanced images , the first is used as the input image to be filtered, and the second is used as the guidance image;

[0030] ;

[0031] In the formula: is the pixel value of the image enhanced by Retinex at (x, y); is the pixel value of the original eye image at (x, y); is the Gaussian function used to simulate the light distribution; represents the convolution operation.

[0032] Furthermore, the Gaussian function has the form of:

[0033] ;

[0034] In the formula: is the standard deviation; is the mean value of the Gaussian function in the x and y directions;

[0035] wherein, is the natural constant, taking 2.71828, , M and N represent the width and height of the user's eye movement video frame.

[0036] Furthermore, after receiving the user's eye movement video frame, the recognition module converts the user's eye movement video frame into a grayscale image, sets the user's pupil grayscale value range, intercepts the local grayscale image representing the user's pupil in the grayscale image based on the user's pupil grayscale value range, and uses the central pixel of the local grayscale image as the reference point;

[0037] All the picked reference points are placed in any user's eye movement video frame for representation;

[0038] When constructing the real eye movement path, the construction module uses the user's eye movement video frame representing all the reference points as the processing target, connects the adjacent reference points in the user's eye movement video frame, and the connection result is the real eye movement path.

[0039] In a second aspect, an intermittent exotropia dynamic monitoring method based on upper and lower frame judgment includes the following steps:

[0040] Collect the user's eye movement image through an eye movement image acquisition device;

[0041] Decompose the user's eye movement image into user's eye movement video frames, set the user's eye movement video frame selection logic, and select the user's eye movement video frames based on the selection logic;

[0042] Obtain the selected user's eye movement video frames, perform optimization processing on the user's eye movement video frames, and create a data packet based on the processed user's eye movement video frames;

[0043] Using the data packet as the processing target, pick reference points in each user's eye movement video frame in the data packet, and construct the user's real eye movement path based on the reference points;

[0044] Comprehensively analyze the similarity between the user's real eye movement path and the predetermined path, set the eye position defect determination threshold, and compare the eye position defect determination threshold with the similarity analysis result to determine whether there is an exotropia defect in the user's eye position.

[0045] Adopting the technical solution provided by the present invention, compared with the known public technology, it has the following beneficial effects:

[0046] The present invention provides an intermittent exotropia dynamic monitoring system and method based on upper and lower frame judgment. During the execution of the system and method, user eye movement frames are captured from the user's eye movement images. Taking the user eye movement frames as basic parameters, the planned eye movement path and the actual eye movement path of the user are obtained. Thus, based on the similarity comparison of the two types of eye movement paths, it is determined whether there is an exotropia defect in the user's eye position, maximizing the application of eye movement monitoring technology to assist medical staff in diagnosing the user's eye position exotropia problem more quickly and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0048] Figure 1 It is a schematic structural diagram of an intermittent exotropia dynamic monitoring system based on upper and lower frame judgment;

[0049] Figure 2 It is a schematic flowchart of an intermittent exotropia dynamic monitoring method based on upper and lower frame judgment;

[0050] Figure 3 It is a schematic diagram of an example of a user eye movement frame in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0052] The following further describes the present invention with reference to the embodiments.

[0053] Embodiment 1:

[0054] An intermittent exotropia dynamic monitoring system based on upper and lower frame judgment in this embodiment, as Figure 1 shown, includes:

[0055] An upload module, used to upload the user's eye movement image and decompose the user's eye movement image into user eye movement frames;

[0056] A selection module, configured to traverse the user eye movement video frames obtained by decomposing in the upload module, and select no less than three user eye movement video frames on each segment of the eye movement path;

[0057] The selection module is provided with sub-modules at a lower level, including:

[0058] An optimization unit, configured to receive the user eye movement video frames selected by the selection module, and perform optimization processing on the user eye movement video frames;

[0059] A sub-packaging unit, configured to receive the user eye movement video frames that have completed the optimization processing in the optimization unit, read the marker time sequence of the user eye movement video frames, determine the source sub-user eye movement image of the user eye movement video frames, and distinguish the user eye movement video frames based on the source sub-user eye movement image of the user eye movement video frames, to obtain a number of user eye movement video frame data packets;

[0060] Among them, the optimization unit is provided with user eye movement video frame optimization processing logic, and the optimization unit performs optimization processing on the user eye movement video frames based on the user eye movement video frame optimization processing logic. The user eye movement video frames included in the user eye movement video frame data packets are sorted and placed based on their respective marker time sequences;

[0061] The optimization processing logic of the user eye movement video frames is expressed as:

[0062] ;

[0063] In the formula: is the pixel value at (x, y) of the finally optimized user eye movement video frame; represents the histogram equalization operation; represents the guided filter operation; is the gamma coefficient;

[0064] Among them, in the guided filter operation stage, both inputs are the enhanced images , the first is used as the input image to be filtered, and the second is used as the guidance image;

[0065] ;

[0066] In the formula: is the pixel value at (x, y) of the image enhanced by Retinex; is the pixel value at (x, y) of the original eye image; is the Gaussian function used to simulate the light distribution; represents the convolution operation;

[0067] Gaussian function has the form of:

[0068] ;

[0069] In the formula: is the standard deviation; is the mean value of the Gaussian function in the x and y directions;

[0070] Among them, is the natural constant, taking 2.71828, , M and N represent the width and height of the user's eye movement video frame;

[0071] Through the above logical formula, the user's eye movement video frame is optimized, effectively improving the accuracy of the system when further running the application to analyze and determine the exotropia of the eye position using the user's eye movement video frame;

[0072] The recognition module is used to receive the user's eye movement video frame selected by the selection module, pick up the reference points in the user's eye movement video frame, and recognize the coordinates of the reference points in each user's eye movement video frame;

[0073] After receiving the user's eye movement video frame, the recognition module converts the user's eye movement video frame into a grayscale image, sets the user's pupil grayscale value range, intercepts the local grayscale image representing the user's pupil in the grayscale image based on the user's pupil grayscale value range, and uses the central pixel of the local grayscale image as the reference point;

[0074] All the picked reference points are placed in any user's eye movement video frame for representation;

[0075] When constructing the real eye movement path, the construction module uses the user's eye movement video frame representing all the reference points as the processing target, connects the adjacent reference points in the user's eye movement video frame, and the connection result is the real eye movement path;

[0076] The construction module is used to receive the coordinates of the reference points recognized by the recognition module and construct the real eye movement path based on the coordinates of the reference points;

[0077] The comparison module is used to compare the similarity between each corresponding real eye movement path and the eye movement path;

[0078] During the operation stage of the comparison module, identify the marking time sequence of the user's eye movement video frame from which the end point of the real eye movement path comes, determine a time domain based on the two marking time sequences, and intercept the local eye movement path in the same time domain in the real eye movement path corresponding to the eye movement path. Further:

[0079] ;

[0080] In the formula: is the comprehensive similarity between each real eye movement path and the eye movement path; is the total amount of real eye movement paths; is the length and slope of the i-th true eye movement path; is the length and slope of the eye movement path corresponding to the i-th true eye movement path; is the adjustment factor;

[0081] Among them, the adjustment factor takes a value of 1 or -1. When the numerator of the fraction where the adjustment factor is located is less than or equal to the denominator, the adjustment factor takes a value of 1. Otherwise, the adjustment factor takes a value of -1;

[0082] Through the above similarity calculation formula, it provides judgment support for the judgment module to judge whether the user's eye position has exotropia.

[0083] The judgment module is used to set the exotropia judgment threshold of the eye position, obtain the similarity obtained by comparison in the comparison module, and when the similarity is not less than the exotropia judgment threshold of the eye position, it is judged that the user's eye position has an exotropia defect. Otherwise, it is judged that the user's eye position does not have an exotropia defect;

[0084] The upload module is wirelessly interactively connected to a selection module. The lower level of the selection module is wirelessly interactively connected to an optimization unit and a sub-packaging unit. The selection module is wirelessly interactively connected to an identification module. The identification module is wirelessly interactively connected to the sub-packaging unit. The identification module is wirelessly interactively connected to a construction module, a comparison module, and a judgment module.

[0085] In this embodiment, the uploading module runs to upload the user's eye movement image, decomposes the user's eye movement image into user's eye movement picture frames. The selection module runs later to traverse the user's eye movement picture frames obtained by decomposition in the uploading module, and selects no less than three user's eye movement picture frames on each segment of the eye movement path. The optimization unit synchronously receives the user's eye movement picture frames selected by the selection module and performs optimization processing on the user's eye movement picture frames. The packetizing unit receives in real time the user's eye movement picture frames that have completed the optimization processing in the optimization unit, reads the marking time sequence of the user's eye movement picture frames, determines that the user's eye movement picture frames are sourced from the sub-user's eye movement image, and based on the sub-user's eye movement image from which the user's eye movement picture frames are sourced, differentiates the user's eye movement picture frames to obtain several user's eye movement picture frame data packets. The recognition module further receives the user's eye movement picture frames selected by the selection module, picks up reference points in the user's eye movement picture frames, identifies the coordinates of the reference points in each user's eye movement picture frame, and then the construction module receives the coordinates of the reference points recognized by the recognition module running, constructs the real eye movement path based on the coordinates of the reference points, and the comparison module compares the similarity between each corresponding real eye movement path and the eye movement path. Finally, the determination module sets the determination threshold for exotropia of the eye position, obtains the similarity obtained by the comparison in the comparison module, and when the similarity is not less than the determination threshold for exotropia of the eye position, determines that the user has an exotropia defect in the eye position, otherwise, determines that the user does not have an exotropia defect in the eye position.

[0086] Through the system in the above embodiment, the eye movement monitoring technology is fully utilized to provide support for the diagnosis of exotropia disease of the eye position, assisting medical staff to make a diagnosis of the patient's exotropia disease of the eye position more quickly and accurately.

[0087] Embodiment 2:

[0088] At the specific implementation level, on the basis of Embodiment 1, this embodiment refers to Figure 1 to further specifically describe a dynamic monitoring system for intermittent exotropia based on upper and lower frame judgment in Embodiment 1:

[0089] The user's eye movement image uploaded by the uploading module is sourced from an eye movement image acquisition device. The eye movement image acquisition device is integrated by a fixed camera, a marker, and an electric slide rail. The marker is installed on the electric slide rail, and the movement of the marker is controlled by the electric slide rail. During the user's eye movement image acquisition stage, the user is prompted to always fixate on the marker, and the marker moves along a predetermined path by the electric slide rail. The fixed camera completes the acquisition of the user's eye movement image during the process of the electric slide rail controlling the movement of the marker and uploads it to the uploading module for storage;

[0090] Among them, there are no less than three bending nodes in the predetermined path used by the electric slide rail to control the movement of the marker, and the frame rate of the fixed camera for acquiring the user's eye movement image is not less than 60 Hz;

[0091] Each user eye movement video frame obtained by decomposition in the upload module is marked with its time sequence in the user eye movement video. The eye movement path is a predetermined path used to control the movement of the marker on the electric slide rail. The user eye movement video is segmented based on the timestamps of the start and end points of each segment of the eye movement path, resulting in several sub-user eye movement videos. Each sub-user eye movement video corresponds to a set of user eye movement video frames, and each set of user eye movement video frames is used as the target for the selection module to select user eye movement video frames, and the selection of user eye movement video frames is performed.

[0092] After the user eye movement video frames selected by the selection module are sorted based on the marked time sequence, the number of intermediate frames between adjacent user eye movement video frames is equal, and the first and last frames among the sorted user eye movement video frames correspond to the start and end timestamps of the source sub-user eye movement video of the frames respectively.

[0093] Among them, the higher the diagnostic accuracy requirement for exotropia of the user's eye position, the more bending nodes there are in the predetermined path and the more user eye movement video frames are selected by the selection module. Conversely, the fewer there are.

[0094] In this embodiment, through the above settings, the operation logic of the upload module is defined, providing the necessary operation data support for the operation of subsequent modules in the system of this embodiment.

[0095] Embodiment 3:

[0096] At the specific implementation level, on the basis of Embodiment 1, this embodiment further specifically describes a dynamic monitoring system for intermittent exotropia based on upper and lower frame judgment in Embodiment 1 with reference to Figure 2 A dynamic monitoring method for intermittent exotropia based on upper and lower frame judgment includes the following steps:

[0097] Collect the user eye movement video through an eye movement video acquisition device;

[0098] Decompose the user eye movement video into user eye movement video frames, set the user eye movement video frame selection logic, and select user eye movement video frames based on the selection logic;

[0099] Obtain the selected user eye movement video frames, perform optimization processing on the user eye movement video frames, and create a data packet based on the processed user eye movement video frames of the user;

[0100] Using the data packet as the processing target, pick reference points in each user eye movement video frame in the data packet, and construct the user's real eye movement path based on the reference points;

[0101] Taking the data packet as the processing target, pick reference points in each user eye movement video frame in the data packet, and construct the user's real eye movement path based on the reference points;

[0102] Comprehensively analyze the similarity between the user's actual eye movement path and the predetermined path, set the eye position defect determination threshold, and determine whether there is an exotropia defect in the user's eye position based on the comparison between the eye position defect determination threshold and the similarity analysis result.

[0103] In summary, during the execution of the system and method in the above embodiments, by capturing the user's eye movement frame in the user's eye movement image and using the user's eye movement frame as a basic parameter, the planned eye movement path and the user's actual eye movement path are obtained. Thus, based on the similarity comparison of the two types of eye movement paths, it is determined whether there is an exotropia defect in the user's eye position, maximizing the application of eye movement monitoring technology to assist medical staff in diagnosing the user's eye position exotropia problem more quickly and accurately.

[0104] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A dynamic monitoring system for intermittent exotropia based on upper and lower frame judgment, characterized in that: include: An uploading module, used for uploading the user's eye movement images and decomposing the user's eye movement images into user's eye animation frames; A selection module is used to traverse the user eye movement image frames decomposed in the upload module, and select no less than three user eye movement image frames on each segment of the eye movement path; A recognition module, used for receiving the user eye animation video frame selected by the selection module, picking up the reference point in the user eye animation video frame, and identifying the coordinates of the reference point in each user eye animation video frame; A construction module, used for receiving the reference point coordinates identified by the recognition module, and constructing a real eye movement path based on the reference point coordinates; A comparison module is used to compare the similarity between the real eye movement path and the eye movement path corresponding to each segment; The determination module is used to set an eye exotropia determination threshold, obtain the similarity obtained by the comparison in the comparison module, and when the similarity is not less than the eye exotropia determination threshold, determine that the user's eye position has an exotropia defect; otherwise, determine that the user's eye position does not have an exotropia defect.

2. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 1 is characterized in that: The user eye movement images uploaded in the upload module are from an eye movement image acquisition device, which is integrated by a fixed camera, a marker, and an electric slide rail. The marker is installed on the electric slide rail, and the movement of the marker is controlled by the electric slide rail. During the user eye movement image acquisition stage, the user is prompted to always look at the marker. The marker moves along a predetermined path through the electric slide rail. The fixed camera completes the acquisition of the user's eye movement images during the process of the electric slide rail controlling the movement of the marker, and uploads them to the upload module for storage; Among them, there are no less than three bending nodes in the predetermined path used by the electric slide rail to control the movement of the marker, and the frame rate of the fixed camera to collect the user's eye movement image is not less than 60Hz.

3. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 2 is characterized in that: The user eye movement image frames decomposed in the upload module all carry their timing marks in the user eye movement image, the eye movement path is the predetermined path used by the electric slide rail to control the movement of the marker, the user eye movement image is segmented based on the departure and arrival timestamps of the endpoints of each segment of the eye movement path, and a plurality of sub-user eye movement images are obtained, each of which corresponds to a set of user eye movement image frames, and each set of user eye movement image frames serves as a target for the selection module to select a user eye movement image frame, and the selection of the user eye movement image frame is performed; After the user eye movement video frames selected by the selection module are sorted based on the marked timing, the number of interval frames between adjacent user eye movement video frames is equal, and the first picture frame and the last picture frame in the sorted user eye movement video frames respectively correspond to the start timestamp and the end timestamp of the eye movement image of the sub-user from which the picture frames are sourced; The higher the requirement for the user's eye exotropia diagnosis accuracy, the more bending nodes of the predetermined path and the more user eye animation frames selected by the selection module, and vice versa.

4. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 3 is characterized in that: The selection module is provided with submodules at the lower level, including: An optimization unit, configured to receive the user eye animation frame selected by the selection module, and optimize the user eye animation frame; A subpacketizing unit is used to receive the user eye animation video frames optimized in the optimization unit, read the marking timing of the user eye animation video frames, determine the eye movement images of the sub-users from which the user eye animation video frames are derived, distinguish the user eye animation video frames based on the eye movement images of the sub-users from which the user eye animation video frames are derived, and obtain a plurality of user eye animation video frame data packets; Among them, the optimization unit is provided with a user eye animation image frame optimization processing logic, and the optimization unit optimizes the user eye animation image frame based on the user eye animation image frame optimization processing logic. The user eye animation image frames included in the user eye animation image frame data packet are sorted and placed based on their respective marking timing.

5. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 4 is characterized in that: The optimization processing logic of the user eye animation frame is expressed as: ； Where: is the pixel value at (x, y) of the final optimized user eye animation frame; Represents the histogram equalization operation; represents the guided filtering operation; is the gamma coefficient; In the guided filtering operation stage, both inputs are enhanced images. , the first As the input image to be filtered, the second as a guiding image; ； Where: is the pixel value at (x, y) of the image after Retinex enhancement; is the pixel value of the original eye image at (x, y); is the Gaussian function used to simulate the illumination distribution; Represents a convolution operation.

6. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 5, characterized in that: The Gaussian function The form is: ； Where: is the standard deviation; is the mean of the Gaussian function in the x and y directions; in, is a natural constant, take 2.71828, , M and N represent the width and height of the user's eye animation frame.

7. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 1 is characterized in that: After receiving the user eye animation graphic frame, the recognition module converts the user eye animation graphic frame into a grayscale image, sets the user pupil grayscale value interval, intercepts a local grayscale image representing the user pupil in the grayscale image based on the user pupil grayscale value interval, and uses the central pixel of the local grayscale image as a reference point; All picked reference points are placed in any user eye animation frame for representation; When constructing the real eye movement path, the construction module takes the user eye animation video frame representing all reference points as the processing target, connects the adjacent reference points in the user eye animation video frame, and the connection result is the real eye movement path.

8. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 1 is characterized in that: During the operation phase of the comparison module, the marking timing of the user's eye animation frame of the real eye movement path endpoint source is identified, a time domain is determined based on two marking timings, and a local eye movement path of the same time domain is intercepted in the determined time domain in the eye movement path corresponding to the real eye movement path, and further: ； Where: is the comprehensive similarity between each real eye movement path and the eye movement path; is the total amount of real eye movement paths; is the length and slope of the i-th real eye movement path; is the length and slope of the eye movement path corresponding to the i-th real eye movement path; is the adjustment factor; Among them, the adjustment factor The value is 1 or -1, adjustment factor The numerator of the fraction is less than or equal to the denominator, and the adjustment factor The value is 1, otherwise, the adjustment factor The value is -1.

9. The intermittent exotropia dynamic monitoring system based on upper and lower frame judgment according to claim 1 is characterized in that: The upload module is interactively connected to the selection module via a wireless network, the selection module is interactively connected to the optimization unit and the subcontracting unit via a wireless network, the selection module is interactively connected to the identification module via a wireless network, the identification module is interactively connected to the subcontracting unit via a wireless network, and the identification module is interactively connected to the construction module, the comparison module and the determination module via a wireless network.

10. A method for dynamic monitoring of intermittent exotropia based on upper and lower frame judgment, the method being an implementation method of a dynamic monitoring system for intermittent exotropia based on upper and lower frame judgment as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: Collecting user's eye movement images through an eye movement image collection device; Decomposing the user eye movement image into user eye movement image frames, setting the user eye movement image frame selection logic, and selecting the user eye movement image frame based on the selection logic; Obtaining the selected user eye animation video frame, optimizing the user eye animation video frame, and creating a data packet based on the user eye animation video frame after the user processing; Taking the data packet as the processing target, picking up reference points in each user eye movement video frame in the data packet, and constructing the user's real eye movement path based on the reference points; Comprehensively analyze the similarity between the user's actual eye movement path and the predetermined path, set the eye position defect judgment threshold, and determine whether the user's eye position has an exotropia defect based on the comparison between the eye position defect judgment threshold and the similarity analysis result.

Images