Full-process intelligent strabismus degree inspection system and method

Abstract

A full-process intelligent strabismus degree inspection system belongs to the field of strabismus degree measurement and comprises an optical display system, a head-mounted strabismus degree detection system and an intelligent control system, and the optical display system comprises a near-vision display system and a far-vision display system; the head-mounted strabismus degree detection system comprises a data set labeling module, a pupil detection positioning module, a sight line deviation angle calculation module, a head posture detection module and a gaze estimation module. The intelligent control system comprises a binocular split vision module, a covering module, an analog prism module, a user input module and a man-machine interaction and data management module, and the optical display system and the head-mounted strabismus degree detection system are both connected with the intelligent control system. The invention further provides a full-process intelligent strabismus degree inspection method. According to the invention, the detection of the conventional strabismus specialized examination can be fully automatically and efficiently realized, the threshold and the cost of the strabismus specialized examination are reduced, and the precision and the stability of the strabismus degree measurement result are greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of strabismus measurement, and in particular to a full-process intelligent strabismus inspection system and method. Background Art

[0002] Strabismus is a common eye disease that affects about 4% of the population. It can be caused by muscle imbalance, neurological problems or genetic factors, which not only affects appearance but can also cause vision problems such as amblyopia or diplopia. Strabismus not only has serious consequences on vision, but also has a serious impact on the patient's mental health, social life, employment, and quality of life. Treatments include glasses, vision training and surgery, and early diagnosis and treatment are very important.

[0003] Since the causes of strabismus are complex, the diagnostic methods for strabismus are also very diverse. Eye movement function tests include eye position tests, diagnostic eye position strabismus tests, extraocular muscle function tests, and eye movement traction tests. Eye position tests include cover-uncover method, alternating cover method, prism alternating cover method, and corneal light reflection method. Diagnostic eye position strabismus tests are performed at nine eye positions. Binocular vision function tests include Worth four-point light test, Bagolini linear mirror test, and stereoscopic vision test. In the above-mentioned specialized examinations for strabismus, the patient's nearsightedness and farsightedness need to be examined, and then the doctor will diagnose and treat based on the results and professional knowledge.

[0004] Clinical eye position examinations often require the participation of professionally trained optometrists, and the measurement time is long and the results are relatively subjective. The gold standard for traditional strabismus measurement is to use a prism, which is placed in front of the patient's eyes to accurately shift the observation image of the strabismic eye back to the normal position. However, a slight tilt of the prism will result in inaccurate image offset, so the prism plane and the line of sight must be kept vertical during use, which places high demands on the optometrist's operation. Clinical binocular vision function examinations require the preparation of corresponding tools for different examination contents, and a sufficiently large venue is required for myopia and farsightedness examinations. Therefore, in clinical practice, comprehensive specialist examinations of strabismus often require a large amount of manpower, material resources and time costs, and it is difficult to obtain accurate, objective and stable measurement results.

[0005] In order to solve the above problems, a Chinese patent with Chinese patent publication number CN117617889A discloses a device and method for measuring the degree of strabismus of the human eye, including a head-supported light source, a voice announcer and a control device. When working, it is not affected by the kappa angle and can accurately measure the degree of strabismus, which solves the shortcomings of the traditional strabismus degree detection technology being relatively subjective and requiring professional doctors to participate in the measurement. However, this method calculates the pupil deflection angle based on Purkin's spots, so there are problems such as dependence on calibration, limited accuracy and sensitivity to reflections when wearing glasses. At the same time, this method still has the problems of difficulty in making precise measurements using prisms and inability to measure farsighted strabismus.

[0006] For example, a Chinese patent with the Chinese patent publication number CN113440095A discloses a device and method for measuring the degree of strabismus of human eyes, including an optical system and a processing control system, and provides a method for measuring the degree of strabismus of human eyes. During operation, a point light source is used as a sight mark, and the angle of pupil deviation is monitored by a camera, and the degree of strabismus is calculated based on the covering method and the prism covering method. However, the hyperopia mark placed by this method through the hyperopia optical path is not completely consistent with the hyperopia mark required for actual measurement, and the accommodative fixation sight mark is not used as the myopia mark, so the measured degree of strabismus may be inaccurate. This method uses a camera inside the device to detect the deviation of the pupil position. Such a design will produce large errors due to slight movements of the patient's head during the measurement process.

[0007] For example, a Chinese patent with the publication number CN105748032A discloses a latent strabismus detection and training instrument, which includes a mechanical unit and an electronic unit. When working, the red-green cross method is used to measure latent strabismus. The patient moves the red and green sight marks on the screen to make them coincide, and then calculates the strabismus degree according to the actual deviation distance of the red and green sight marks. However, the red-green cross method is a method for measuring subjective strabismus, which requires the patient's understanding and cooperation, and is therefore not suitable for young children and those who are not cooperative.

[0008] At present, there is no method at home and abroad that can completely replace professional doctors in conducting comprehensive strabismus specialist examinations, including integrated myopia and hyperopia optical paths, high-precision pupil deflection angle detection for strabismus measurement, implementation of cover method eye position examination, overcoming the difficulty of measuring strabismus with prisms and realizing binocular vision function examination. Summary of the invention

[0009] In order to overcome the single means of strabismus measurement in existing strabismus measurement systems and methods, which cannot meet the needs of strabismus diagnosis and treatment in various situations, the present invention provides a full-process intelligent strabismus inspection system and method, which uses a high-precision strabismus measurement system based on line of sight estimation technology, integrates the near-sighted and far-sighted optical paths, and overcomes the difficulty of measuring physical prisms by simulating prism methods, so that the conventional strabismus specialist inspection method can be fully automatically and efficiently implemented. The present invention reduces the threshold and cost of strabismus specialist inspections and greatly improves the accuracy and stability of strabismus measurement results.

[0010] The technical solution of the present invention is as follows: A full-process intelligent strabismus inspection system comprises an optical display system, a head-mounted strabismus detection system and an intelligent control system, wherein the optical display system comprises a near vision display system and a far vision display system; the head-mounted strabismus detection system comprises a data set annotation module, a pupil detection and positioning module, a sight line deviation angle calculation module, a head posture detection module and a gaze estimation module; the intelligent control system comprises a binocular dichoptic module, a covering module, a simulated prism module, a user input module and a human-computer interaction and data management module, and the optical display system and the head-mounted strabismus detection system are both connected to the intelligent control system.

[0011] Furthermore, the optical display system includes a near vision display system and a far vision display system, wherein: The near vision display system includes a near vision display screen, a near vision sight mark light source and a display spectroscope. The near vision display screen and the near vision sight mark are used for adjusting the fixation sight mark and the point light source sight mark respectively. The near vision display screen and the near vision sight mark are imaged near the eyes of the person being measured through the display spectroscope. The telescopic display system includes a telescopic display screen, a telescopic sight mark light source, a display beam splitter and a concave reflector. The telescopic display screen and the telescopic sight mark are used for the adjustable fixation sight mark and the point light source sight mark respectively. The telescopic display screen and the telescopic sight mark are reflected by the display beam splitter into the concave reflector, reflected from the concave reflector, and finally formed through the display beam splitter at a distance in front of the eyes of the person being measured.

[0012] The accommodative fixation sight mark refers to a sight mark with fine details (such as patterns, letters, etc.).

[0013] Furthermore, the head-mounted strabismus detection system includes a data set annotation module, a pupil detection and positioning module, a sight line deviation angle calculation module, a head posture detection module and a gaze estimation module, wherein: The dataset annotation module is used to collect and annotate eye image datasets for training deep learning models; Pupil detection and positioning module, used to locate the center of the pupil; The sight line deviation angle calculation module is used to collect the eye images of the human eye and calculate the difference of the sight line angle of the human eye at two different moments according to the eye images of the human eye at two different moments; A head posture detection module, used to detect the head posture of the subject; The gaze estimation module is used to detect the gaze position of a single eye of the subject on the display screen.

[0014] Preferably, the implementation process of the dataset annotation module is as follows: Step 111, recruiting data collection subjects, selecting a plurality of persons with normal visual function as data collection subjects; Step 112, data collection, the subject wears a head-mounted strabismus detection device, and the near vision display screen randomly displays an accommodative fixation sight mark. After the subject's line of sight is stable, the eye images of both eyes and the coordinates of the sight mark are saved at this moment; Step 113, prepare the ellipse equation and use the ellipse fitting method to mark the pupil. The standard ellipse equation is ; Step 114, marking pupil edge points, manually marking more than five adjacent points of the pupil and iris; Step 115, construct a matrix equation, substitute each point into the ellipse equation, and obtain a linear equation system; Step 116, solving the problem, using the least square method to solve the overdetermined system of equations, by singular value decomposition or directly calling the numerical calculation library to solve; Step 117, verify the result: draw an ellipse using the calculated parameters A, B, C, D, E, and F, and compare it with the original data points to verify the fitting effect; Step 118, calculate the center point of the ellipse and save it as the center point of the pupil and the anchor box.

[0015] More preferably, the implementation process of the pupil detection and positioning module is as follows: Step 121, data preparation, using a data set annotation module to prepare a sufficient number of data sets; Step 122, data preprocessing, normalizing and scaling the image for data enhancement to increase the robustness of the model; Step 123, model selection, select a suitable neural network architecture, such as U-Net, ResNet and YOLO; Step 124, model training, using the labeled data set to train the model, with the input being the eye image and the output being the pupil anchor box; Step 125, model optimization, adjusts hyperparameters such as network structure, learning rate, batch size, etc. to optimize the model recognition accuracy.

[0016] More preferably, the implementation process of the sight line deviation angle calculation module is as follows: Step 131, using the pupil detection and positioning module to detect the captured eye image in real time, and saving the images at time t0 and time t1 when the pupil is successfully identified; Step 132, using the pupil detection and positioning module to calculate the center position x0 and x1 of the pupil at time t0 and time t1; Step 133, offset angle , where θ0 is the angle between the measured eye's sight line and the vertical line of the fixation plane at time t0, , R is the radius of the eyeball; Furthermore, the eyeball radius is obtained by collecting images of normal human eyes at known deviation angles, calculating Δx between images, and fitting to obtain the eyeball radius parameter R. The eyeball radius R of an average adult is about 12.5 mm in the world coordinate system. The eyeball radius parameter R of different age groups with statistical significance can be loaded during device initialization, and the human eye model can also be calibrated for the person being tested.

[0017] Even more preferably, the implementation process of the head posture detection module is as follows: Step 141, acquiring data, acquiring real-time (X, Y, Z) angular velocity information from the head posture sensor; Step 142, data processing, integrating the angular velocity data to obtain the angle change of the head, and using a filtering algorithm, such as Kalman filtering or complementary filtering, to reduce the influence of noise; Step 143, posture estimation, using the integrated angle change to estimate the posture (pitch, roll, yaw) of the head, and performing posture correction in combination with the accelerometer data to reduce drift error; Step 144, implement posture detection, set a threshold or pattern recognition algorithm according to the posture change, and detect specific head movements.

[0018] The intelligent control system includes a binocular vision module, a covering module, a simulated prism module, a user input module, and a human-computer interaction and data management module, wherein: Binocular separation module, used to separate the images observed by the person under test, so that when the person under test observes the same display plane, both eyes can obtain different images; A covering module is used to cover the observation eye of the person being tested; The simulated prism module is used to replace the traditional prism to achieve the function of accurately shifting the measured person's strabismus to observe the picture, and to achieve accurate measurement of the measured person's strabismus; A user input module is used to assist the person being measured to move the sight mark in the image to achieve the measurement of subjective strabismus; The human-computer interaction and data management module is used to assist testers in using the equipment and to process, store and print data after the measurement.

[0019] Furthermore, the binocular vision separation technology may be polarization vision separation, red and blue vision separation, liquid crystal shutter vision separation, etc.

[0020] Furthermore, the covering module is composed of a liquid crystal light valve module and a binocular dichoptic module, and the implementation process of the covering module is as follows: Step 211, selecting the occluded eye; Step 212, closing the display image of the blocked eye in the binocular dichoptic module; Step 213, closing the liquid crystal light valve module that blocks the eyes.

[0021] Furthermore, the liquid crystal light valve module includes a polarizing layer, a liquid crystal layer and a polarizing layer. The blocking of the liquid crystal light valve module will not affect the camera to capture the eye image. The implementation process of the liquid crystal light valve module is as follows: Step 221, keeping the polarization directions of the polarizing layer and the analyzing layer consistent; Step 222, when no electric field is applied to the liquid crystal layer, the light of the display entering from the polarizing layer does not change its polarization direction after passing through the liquid crystal layer, and can pass through the analyzer layer; Step 223, when the controller applies an electric field to the liquid crystal layer, the light entering from the polarizing layer has a polarization direction perpendicular to the original direction after passing through the liquid crystal layer, and cannot pass through the analyzing layer, thereby achieving a shielding effect; Step 224, after the light from the eye of the person under test passes through the polarization analyzer layer and the liquid crystal layer, it is directly reflected into the camera by the camera beam splitter, so the camera can capture the image regardless of whether the liquid crystal layer applies an electric field.

[0022] The implementation process of the simulated prism module is as follows: Step 231: selecting a near vision display screen or a far vision display screen as a display screen for simulating a prism module, and the distance between the virtual image of the display screen after passing through the optical path of the optical display system and the eyes of the person being measured is d; Step 232: using a binocular vision module to separate the images observed by the left and right eyes of the person under test, so that the left eye of the person under test can only observe the left eye observation image of the screen, and the right eye can only observe the right eye observation image of the screen; Step 233: By shifting the position of the strabismus eye observing the screen, the observation effect of the person under test using a prism is simulated. Assuming that the degree of the prism to be placed is △, the separation of the sight mark is L, and the angle between the main eye's line of sight and the perpendicular line of the display plane is θ, then , where L is in cm and d is in m.

[0023] The user input module is composed of input units such as a mouse, a keyboard or a handle, which is convenient for the person being tested to control the position of the sight mark in the observation picture.

[0024] A full-process intelligent strabismus examination method, including an eye movement function examination process and a binocular vision function examination process; The eye movement function test process includes the cover-uncover method, the alternating cover method, the prism cover method, the simulated corneal light reflection method and the nine eye position test method; The binocular vision function test process includes Worth four-point light test method, Bagolini linear mirror test method and stereoscopic test method.

[0025] Furthermore, the cover-uncover method is used to measure the strabismus of a single eye. The implementation process of the cover-uncover method is as follows: Step 311, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 312, displaying an accommodative fixation target on the near vision display screen, and recording the image I_0 of the eye being tested after the person's line of sight is stabilized; Step 313, using a covering module to cover the non-tested eye on the opposite side of the tested eye, and recording the image I_1 of the tested eye at this moment after the tested person's line of sight is stabilized; Step 314, using a sight deviation angle calculation module to calculate the gaze deviation angle of the measured eye from image I_0 to image I_1; Step 315, repeating the measurement process from step 312 to step 313 for multiple times, and calculating the average of the multiple results, to obtain the average gaze deviation angle of the measured eye as the myopia degree of the measured eye; Step 316, if the measured eye does not move or the strabismus is too small, measure the other eye in the same way; Step 317, changing the near vision display screen in step 312 to a far vision display screen or a far vision sight mark light source, repeating steps 312 to 316, and obtaining the far vision strabismus of the eye being tested.

[0026] Further, the implementation process of the alternating covering method is as follows: Step 321, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 322, displaying an accommodative fixation sight mark on the near vision display screen, and waiting for the line of sight of the person being measured to stabilize; Step 323, using a covering module to cover the left eye, waiting for the person being tested to stabilize his / her sight line, and then recording the left eye image I_left_0 and the right eye image I_right_0 at this moment; Step 324, using the covering module to cover the right eye, waiting for the measured person's sight to stabilize, and then recording the left eye image I_left _1 and the right eye image I_right _1 at this moment; Step 325, using the sight deviation angle calculation module to calculate the gaze deviation angle of the measured eye from image I_left_1 to image I_left_0, recorded as the strabismus of the left eye this time; Step 326, using the sight deviation angle calculation module to calculate the gaze deviation angle of the eye under test from image I_right_0 to image I_right_1, recorded as the strabismus of the right eye this time; Step 327, repeating the measurement process from step 322 to step 326 for multiple times, and calculating the average of the multiple results, to obtain the average gaze deviation angle of the left and right eyes as the strabismus of the measured eye; Step 328, changing the near vision display screen in step 322 to a far vision display screen or a far vision sight mark light source, repeating steps 322 to 327, and obtaining the far vision strabismus of the left and right eyes.

[0027] Further, the implementation process of the prism covering method is as follows: Step 331, initialize array D i Used to store the measured strabismus, where i is the number of times the cover-uncover method or the alternating cover method is used after adding the prism, i = 0, 1, 2, 3 ...; set D0 to the strabismus measured without adding the prism or the strabismus input by the user, and set the prism degree D at this moment to D0; Step 332, implement the cover-uncover method or the alternating cover method to measure the strabismus degree D of the person under test under the prism. i ; Step 333, if the strabismus degree D is measured i is 0, then the strabismus of the person being measured at this moment is If the measured strabismus degree D i If it is not 0, then set the prism degree D at this moment to Then repeat step 332.

[0028] Preferably, in step 331, the prism degree may be set by placing a prism lens on a replaceable lens holder or by using a simulated prism module.

[0029] Furthermore, the implementation process of the simulated corneal light reflection method is as follows: Step 341, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 342, displaying an accommodative fixation sight mark on the near vision display screen, and waiting for the line of sight of the person being measured to stabilize; Step 343, using a gaze estimation module to calculate the gaze positions of the left and right eyes on the screen; Step 344, calculate the degree of strabismus of the left and right eyes according to the deviation between the left and right eyes' gaze positions on the screen and the actual sight mark. Assuming that the distance from the sight mark to the eyes of the person being tested is d, and the distance from the gaze point to the sight mark is L, then the strabismus degree D is m =L / d, where L is in cm and d is in m; Step 345, changing the near vision display screen in step 342 to a far vision display screen or a far vision sight mark light source, repeating steps 342 to 344, and obtaining the far vision strabismus of the left and right eyes.

[0030] Further, the implementation process of the nine-eye position inspection method is as follows: Step 351, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 352, setting nine eye position sight mark positions on the near vision display screen, respectively in the original position where the person being measured is looking, horizontally to the right, horizontally to the left, to the upper left, to the lower left, to the upper right, to the lower right, vertically upward, and vertically downward; Step 353, lighting up the sight mark accordingly, and implementing the cover-uncover method, the alternating cover method, the prism cover method or the simulated corneal light reflection method to measure the myopia degree of the person being measured at the nine eye positions; Step 354, setting a sight mark at the center of the telescopic display screen; Step 355, guide the person being measured to rotate his head in the original position, horizontally to the right, horizontally to the left, to the upper left, to the lower left, to the upper right, to the lower right, vertically upward, and vertically downward directions respectively through the voice announcer. After the head posture detection module confirms that the head posture is correct, implement the cover-uncover method, the alternating cover method, the prism cover method, or the simulated corneal light reflection method to measure the degree of hypertropia of the person being measured in the nine eye positions.

[0031] During the binocular vision function test, the implementation process of the Worth four-light test method is as follows: Step 411, after the person being measured wears the head-mounted strabismus detection device, the person being measured places his head on a head fixing bracket, and before the measurement starts, the person being measured is guided by a voice announcer to observe the image displayed on the display screen; Step 412, lighting the Worth four-point lamp in the near-sighted sight mark light source, wherein the right eye image displays sight mark 1 and sight mark 3, and the left eye image displays sight mark 2, sight mark 3, and sight mark 4; Step 413, allowing the person being tested to select the image that he or she observes, and observing different images corresponds to different visual function problems of the person being tested; Step 414, changing the near vision target light source in step 412 to the far vision target light source, repeating steps 412 and 413 to obtain the far vision measurement result; The implementation process of the Bagolini linear microscopy method is as follows: Step 421, the person being measured wears the head-mounted strabismus detection device, turns off the binocular dichoptic module, installs the Bagolini linear mirror on the replaceable lens holder, places the head on the head fixing bracket, and guides the person being measured to observe the image displayed on the display screen through the voice announcer before the measurement starts; Step 422, lighting a white light source for Bagolini linear speculum examination in the nearsightedness target light source, judging the visual acuity of the person being tested according to the person being tested's feedback, a person with normal binocular single vision function sees a complete cross, a patient with monocular gaze can see a cross when binocularly gazing due to peripheral fusion, but because the macula of the non-dominant eye is suppressed, the center of the stripe seen by the non-dominant eye is discontinuous; Step 423, changing the nearsightedness sight mark light source in step 422 to a farsightedness sight mark, repeating step 422, and obtaining the inspection result under the farsightedness condition; The implementation process of the stereoscopic inspection method is as follows: Step 431, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement starts, the person being measured is guided by a voice announcer to observe the image displayed on the display screen; Step 432, combining with the binocular dichoptic module, displaying near stereoscopic vision inspection images such as Titmus stereogram and TNO stereogram on the near vision display screen to inspect the near stereoscopic vision function of the person being tested; Step 433, in combination with the binocular dichoptic module, a far stereoscopic vision test chart such as a Randot test chart and a Frisby-Davis test chart is displayed on the far vision display screen to check the far stereoscopic vision function of the person being tested.

[0032] The beneficial effects of the present invention are: it can realize conventional strabismus examination fully automatically and efficiently; it reduces the threshold and cost of strabismus specialist examination, and greatly improves the accuracy and stability of strabismus measurement results. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a structural diagram of a full-process intelligent strabismus inspection system; Figure 2 It is a schematic diagram of the execution of a full-process intelligent strabismus inspection method; Figure 3 It is a schematic diagram of the optical path of the optical display system; Figure 4 is a schematic diagram of the structure of a liquid crystal light valve module; Figure 5 It is a flow chart for realizing the prism covering method; Figure 6 It is a schematic diagram of the main part of the equipment of a full-process intelligent strabismus inspection system; Figure 7 It is a full-process intelligent strabismus inspection system. The back diagram of the head-mounted part of the equipment; Figure 8 It is a front view of the head-mounted part of the full-process intelligent strabismus inspection system equipment; Fig. 9 This is a diagram of the Worth four-point light sight mark. DETAILED DESCRIPTION

[0034] The present invention is further described below in conjunction with the accompanying drawings and examples. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent changes of the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0035] Reference Figure 1 , a full-process intelligent strabismus inspection system, including an optical display system, a head-mounted strabismus detection system and an intelligent control system. Figure 6 , Figure 7 and Figure 8 As shown, it is a schematic diagram of a full-process intelligent strabismus inspection system. The optical display system includes a head fixing bracket 107, a first shell 108, a near vision display screen 101, a near vision sight mark light source 102, a far vision display screen 104, a far vision sight mark light source 105, a display spectroscope 103 and a concave reflector 106; the head-mounted strabismus detection system includes a second shell 301, an infrared lighting source 201, a camera spectroscope 205, a polarizing layer 202, a liquid crystal layer 203, a polarizing layer 204, a camera 206, a binocular dichroic device 302, a replaceable lens frame 303, a voice announcer 304 and a head posture sensor 305; the near vision display screen, the near vision sight mark light source, the far vision display screen, the far vision sight mark light source, the camera, the binocular dichroic device, the voice announcer and the head posture sensor are all connected to the intelligent control system.

[0036] The optical display system includes a near vision display system and a far vision display system. Figure 3 It is a schematic diagram of the optical path of the optical display system, where: The near vision display system includes a near vision display screen 101, a near vision sight mark light source 102 and a display spectroscope 103. The near vision display screen and the near vision sight mark are used for the adjustable fixation sight mark and the point light source sight mark, respectively. The near vision display screen and the near vision sight mark are imaged 33 cm in front of the eyes of the measured person through the display spectroscope.

[0037] The telescopic display system includes a telescopic display screen 104, a telescopic sight mark light source 105, a display spectroscope 103 and a concave reflector 106. The telescopic display screen and the telescopic sight mark are used for the adjustable fixation sight mark and the point light source sight mark, respectively. The telescopic display screen and the telescopic sight mark are reflected by the display spectroscope into the concave reflector, reflected from the concave reflector, and finally imaged at 6m in front of the eyes of the measured person through the display spectroscope.

[0038] The accommodative fixation sight mark refers to a sight mark with fine details (such as patterns, letters, etc.).

[0039] The head-mounted strabismus detection system comprises a data set annotation module, a pupil detection and positioning module, a sight line deviation angle calculation module, a head posture detection module and a gaze estimation module, wherein: The dataset annotation module is used to collect and annotate eye image datasets for training deep learning models; Pupil detection and positioning module, used to locate the center of the pupil; The sight line deviation angle calculation module is used to collect the eye images of the human eye and calculate the difference of the sight line angle of the human eye at two different moments according to the eye images of the human eye at two different moments; A head posture detection module, used to detect the head posture of the subject; The gaze estimation module is used to detect the gaze position of a single eye of the subject on the display screen.

[0040] Furthermore, the implementation process of the dataset annotation module is as follows: Step 111, recruiting data collection subjects, selecting a plurality of persons with normal visual function as data collection subjects; Step 112, data collection, the subject wears a head-mounted strabismus detection device, and the near vision display screen randomly displays an accommodative fixation sight mark. After the subject's line of sight is stable, the eye images of both eyes and the coordinates of the sight mark are saved at this moment; Step 113, prepare the ellipse equation and use the ellipse fitting method to mark the pupil. The standard ellipse equation is ; Step 114, marking pupil edge points, manually marking more than five adjacent points of the pupil and iris; Step 115, construct the matrix equation, for each point , substituting it into the elliptic equation, we get the linear equation , write the equations for all points in matrix form ,in: ; Step 116, solving the problem, using the least square method to solve the overdetermined system of equations, using the numpy.linalg.lstsq function in the NumPy library to solve the system of equations; Step 117, verify the result: draw an ellipse using the calculated parameters A, B, C, D, E, and F, and compare it with the original data points to verify the fitting effect; Step 118, calculate the center point of the ellipse and save it as the center point of the pupil and the anchor box.

[0041] Further, the implementation process of the pupil detection and positioning module is as follows: Step 121, data preparation, using a data set annotation module to prepare a sufficient number of data sets; Step 122, data preprocessing, adjusting the image to a size of 640*640 by proportional scaling and zero padding as model input; Step 123, model selection, select the YOLOv8 model as the training model; Step 124, model training, using the labeled data set to train the model, inputting a 640*640 eye image, and outputting a pupil anchor box of (x, y, w, h), where x, y are the coordinates of the anchor box in the x-axis and y-axis directions, and w, h are the width and height of the anchor box; Step 125, model optimization, set the initial learning rate to 0.004, and use the Adam optimizer to dynamically adjust the learning rate during training; set the training batch to 16.

[0042] Furthermore, the implementation process of the sight line deviation angle calculation module is as follows: Step 131, using the pupil detection and positioning module to detect the captured eye image in real time, and saving the images at time t0 and time t1 when the pupil is successfully identified; Step 132, using the pupil detection and positioning module to calculate the center position x0 and x1 of the pupil at time t0 and time t1; Step 133, offset angle , where θ0 is the angle between the measured eye's sight line and the vertical line of the fixation plane at time t0, , R is the radius of the eyeball.

[0043] The eyeball radius is obtained by collecting images of normal human eyes at known deviation angles, calculating Δx between images, and fitting to obtain the eyeball radius parameter R. The eyeball radius R of an average adult is about 12.5 mm in the world coordinate system. You can choose to load the eyeball radius parameter R of different age stages with statistical significance when initializing the device, or you can calibrate the human eye model for the person being tested.

[0044] Further, the implementation process of the head posture detection module is as follows: Step 141, acquiring data, acquiring real-time (X, Y, Z) angular velocity information from the head posture sensor; Step 142, data processing, integrating the angular velocity data to obtain the angle change of the head, and using Kalman filtering to reduce the influence of noise; Step 143, attitude estimation, uses the integrated angle change to estimate the attitude (pitch, roll, yaw) of the head, and performs attitude correction in combination with the accelerometer data to reduce drift errors.

[0045] Step 144, implement posture detection, set a threshold or pattern recognition algorithm according to the posture change, and detect specific head movements.

[0046] The intelligent control system includes a binocular vision module, a covering module, a simulated prism module, a user input module, and a human-computer interaction and data management module, wherein: Binocular separation module, used to separate the images observed by the person under test, so that when the person under test observes the same display plane, both eyes can obtain different images; A covering module is used to cover the observation eye of the person being tested; The simulated prism module is used to replace the traditional prism to achieve the function of accurately shifting the measured person's strabismus to observe the picture, and to achieve accurate measurement of the measured person's strabismus; A user input module is used to assist the person being measured to move the sight mark in the image to achieve the measurement of subjective strabismus; The human-computer interaction and data management module is used to assist testers in using the equipment and to process, store and print data after the measurement.

[0047] Furthermore, the binocular vision separation technology may be polarization vision separation, red and blue vision separation, liquid crystal shutter vision separation, etc.

[0048] Furthermore, the liquid crystal light valve module includes a polarizing layer, a liquid crystal layer and a polarizing layer. The blocking of the liquid crystal light valve module will not affect the camera to capture the eye image. Figure 42 is a schematic diagram of the structure of the liquid crystal light valve module, including an infrared illumination light source 201, a polarizing layer 202, a liquid crystal layer 203, a polarizing layer 204, a camera beam splitter 205, a camera 206, a display 207 and a controller 208. The implementation process of the liquid crystal light valve module is as follows: Step 211, selecting the occluded eye; Step 212, closing the display image of the blocked eye in the binocular dichoptic module; Step 213, closing the liquid crystal light valve module that blocks the eyes.

[0049] Furthermore, the liquid crystal light valve module comprises a polarizing layer, a liquid crystal layer and a polarizing layer. The blocking of the liquid crystal light valve module will not affect the camera to capture the eye image. The implementation process of the liquid crystal light valve module is as follows: Step 221, keeping the polarization directions of the polarizing layer and the analyzing layer consistent; Step 222, when no electric field is applied to the liquid crystal layer, the polarization direction of the light entering from the polarizing layer does not change after passing through the liquid crystal layer, and can pass through the analyzing layer; Step 223, when an electric field is applied to the liquid crystal layer, the light entering from the polarizing layer has a polarization direction perpendicular to the original direction after passing through the liquid crystal layer, and cannot pass through the analyzing layer, thereby achieving a shielding effect; Step 224, after the light from the eye of the person under test passes through the polarization analyzer layer and the liquid crystal layer, it is directly reflected into the camera by the camera beam splitter, so the camera can capture the image regardless of whether the liquid crystal layer applies an electric field.

[0050] Furthermore, the implementation process of the simulated prism module is as follows: Step 231: selecting a near vision display screen or a far vision display screen as a display screen for simulating a prism module, and the distance between the virtual image of the display screen after passing through the optical path of the optical display system and the eyes of the person being measured is d; Step 232: using a binocular vision module to separate the images observed by the left and right eyes of the person under test, so that the left eye of the person under test can only observe the left eye observation image of the screen, and the right eye can only observe the right eye observation image of the screen; Step 233: By shifting the position of the strabismus eye observing the screen, simulate the observation effect of the person under test using a prism. Assuming that the degree of the prism to be placed is △, the separation of the sight mark is L, and the angle between the main eye's line of sight and the perpendicular line of the display plane is θ, then , where L is in cm and d is in m.

[0051] Furthermore, the user input module is composed of input units such as a mouse, a keyboard or a handle, so that the person being tested can control the position of the sight mark in the observation screen.

[0052] A full-process intelligent strabismus examination method includes an eye movement function examination process and a binocular vision function examination process.

[0053] The eye movement function examination process includes the cover-uncover method, the alternating cover method, the prism cover method, the simulated corneal light reflection method and the nine eye position examination method.

[0054] The cover-uncover method can measure the strabismus of a single eye. The implementation process of the cover-uncover method is as follows: Step 311, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 312, displaying an accommodative fixation target on the near vision display screen, and recording the image I_0 of the eye being tested after the person's line of sight is stabilized; Step 313, using a covering module to cover the non-tested eye on the opposite side of the tested eye, and recording the image I_1 of the tested eye at this moment after the tested person's line of sight is stabilized; Step 314, using a sight deviation angle calculation module to calculate the gaze deviation angle of the measured eye from image I_0 to image I_1; Step 315, repeating the measurement process from step 312 to step 313 for multiple times, and calculating the average of the multiple results, to obtain the average gaze deviation angle of the measured eye as the myopia degree of the measured eye; Step 316, if the measured eye does not move or the strabismus is too small, measure the other eye in the same way; Step 317, changing the near vision display screen in step 312 to a far vision display screen or a far vision sight mark light source, repeating steps 312 to 316, and obtaining the far vision strabismus of the eye being tested.

[0055] The implementation process of the alternating covering method is as follows: Step 321, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 322, displaying an accommodative fixation sight mark on the near vision display screen, and waiting for the line of sight of the person being measured to stabilize; Step 323, using a covering module to cover the left eye, waiting for the person being tested to stabilize his / her sight line, and then recording the left eye image I_left_0 and the right eye image I_right_0 at this moment; Step 324, using the covering module to cover the right eye, waiting for the measured person's sight to stabilize, and then recording the left eye image I_left _1 and the right eye image I_right _1 at this moment; Step 325, using the sight deviation angle calculation module to calculate the gaze deviation angle of the measured eye from image I_left_1 to image I_left_0, recorded as the strabismus of the left eye this time; Step 326, using the sight deviation angle calculation module to calculate the gaze deviation angle of the eye under test from image I_right_0 to image I_right_1, recorded as the strabismus of the right eye this time; Step 327, repeating the measurement process from step 322 to step 326 for multiple times, and calculating the average of the multiple results, to obtain the average gaze deviation angle of the left and right eyes as the strabismus of the measured eye; Step 328, changing the near vision display screen in step 322 to a far vision display screen or a far vision sight mark light source, repeating steps 322 to 327, and obtaining the far vision strabismus of the left and right eyes.

[0056] The flow chart of the prism covering method is as follows Figure 5 As shown, the implementation process is as follows: Step 331, initialize array D i Used to store the measured strabismus, where i is the number of times the cover-uncover method or the alternating cover method is used after adding the prism, i = 0, 1, 2, 3 ...; set D0 to the strabismus measured without adding the prism or the strabismus input by the user, and set the prism degree D at this moment to D0; Step 332, implement the cover-uncover method or the alternating cover method to measure the strabismus degree D of the person under test under the prism. i ; Step 333, if the strabismus degree D is measured i is 0, then the strabismus of the person being measured at this moment is If the measured strabismus degree D i If it is not 0, then set the prism degree D at this moment to Then repeat step 332.

[0057] Preferably, in step 331, the prism degree may be set by placing a prism lens on a replaceable lens holder or by using a simulated prism module.

[0058] The implementation process of the simulated corneal light reflection method is as follows: Step 341, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 342, displaying an accommodative fixation sight mark on the near vision display screen, and waiting for the line of sight of the person being measured to stabilize; Step 343, using a gaze estimation module to calculate the gaze positions of the left and right eyes on the screen; Step 344, calculate the degree of strabismus of the left and right eyes according to the deviation between the left and right eyes' gaze positions on the screen and the actual sight mark. Assuming that the distance from the sight mark to the eyes of the person being tested is d, and the distance from the gaze point to the sight mark is L, then the strabismus degree D is m =L / d, where L is in cm and d is in m; Step 345, changing the near vision display screen in step 342 to a far vision display screen or a far vision sight mark light source, repeating steps 342 to 344, and obtaining the far vision strabismus of the left and right eyes.

[0059] The implementation process of the nine-eye position inspection method is as follows: Step 351, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 352, set nine eye position sight mark positions on the near vision display screen, respectively deflected by 25 degrees in the original position where the person being tested is looking, horizontally to the right, horizontally to the left, to the upper left, to the lower left, to the upper right, to the lower right, vertically upward, and vertically downward.

[0060] Step 353, lighting up the sight mark accordingly, and implementing the cover-uncover method, the alternating cover method, the prism cover method or the simulated corneal light reflection method to measure the myopia degree of the person being measured at the nine eye positions; Step 354, setting a sight mark at the center of the telescopic display screen; Step 355, guide the person being measured to rotate his head 25 degrees in the original position, horizontally to the right, horizontally to the left, to the upper left, to the lower left, to the upper right, to the lower right, vertically upward, and vertically downward directions respectively through the voice announcer. After the head posture detection module confirms that the head posture is correct, implement the cover-uncover method, the alternating cover method, the prism cover method or the simulated corneal light reflection method to measure the degree of hypertropia of the person being measured in the nine eye positions.

[0061] The binocular vision function test process includes Worth four-point light test method, Bagolini linear mirror test method and stereoscopic test method.

[0062] The implementation process of the Worth four-light inspection method is as follows: Step 411, after the person being measured wears the head-mounted strabismus detection device, the person being measured places his head on a head fixing bracket, and before the measurement starts, the person being measured is guided by a voice announcer to observe the image displayed on the display screen; Step 412, as shown in the attached Fig. 9 As shown, the Worth four-point lamp is lit in the near sight mark light source, wherein the right eye screen displays sight mark 1 and sight mark 3, and the left eye screen displays sight mark 2, sight mark 3 and sight mark 4; Step 413, allowing the person being tested to select the image that he or she observes, and observing different images corresponds to different visual function problems of the person being tested; Step 414, changing the near vision sight mark light source in step 412 to the far vision sight mark light source, repeating steps 412 and 413, and obtaining the far vision measurement result.

[0063] The implementation process of the Bagolini linear microscopy method is as follows: Step 421, the person being measured wears the head-mounted strabismus detection device, turns off the binocular dichoptic module, installs the Bagolini linear mirror on the replaceable lens holder, places the head on the head fixing bracket, and guides the person being measured to observe the image displayed on the display screen through the voice announcer before the measurement starts; Step 422, light up the white light source for Bagolini linear speculum examination in the nearsightedness target light source, and judge the visual acuity of the person being tested according to the feedback of the person being tested. A person with normal binocular single vision function sees a complete cross. A patient with monocular fixation can see a cross when fixating with both eyes due to peripheral fusion, but because the macula of the non-dominant eye is suppressed, the center of the stripe seen by this eye is discontinuous.

[0064] Step 423, changing the near-sightedness sight mark light source in step 422 to a far-sightedness sight mark, repeating step 422, and obtaining the inspection result under the far-sightedness condition.

[0065] The implementation process of the stereoscopic inspection method is as follows: Step 431, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement starts, the person being measured is guided by a voice announcer to observe the image displayed on the display screen; Step 432, in combination with the binocular dichoptic module, near stereoscopic vision inspection images such as the Titmus stereogram and the TNO stereogram are displayed on the near vision display screen to inspect the near stereoscopic vision function of the person being tested.

[0066] Step 433, in combination with the binocular dichoptic module, a far stereoscopic vision test chart such as a Randot test chart and a Frisby-Davis test chart is displayed on the far vision display screen to check the far stereoscopic vision function of the person being tested.

[0067] The measurement method of this embodiment is to obtain the measured value of strabismus by computer processing and calculation. The measured value does not belong to the medical diagnosis basis for judging whether the patient suffers from strabismus. The measurement method of the present invention is a non-medical technical solution, and its invention purpose is to achieve non-medical strabismus measurement. From the medical diagnosis of whether a patient suffers from strabismus, in addition to strabismus measurement as a reference, it is usually necessary to combine other detection methods and the patient's clinical manifestations for comprehensive diagnosis.

[0068] The contents described in the embodiments of this specification are merely enumerations of implementation forms of the inventive concept and are for illustrative purposes only. The protection scope of the present invention should not be considered to be limited to the specific forms described in this embodiment, and the protection scope of the present invention also extends to equivalent technical means that can be thought of by ordinary technicians in this field based on the inventive concept.

Claims

1. A full-process intelligent strabismus inspection system, characterized in that: The full-process intelligent strabismus inspection system includes an optical display system, a head-mounted strabismus detection system and an intelligent control system. The optical display system includes a near vision display system and a far vision display system; the head-mounted strabismus detection system includes a data set annotation module, a pupil detection and positioning module, a line of sight deviation angle calculation module, a head posture detection module and a gaze estimation module; the intelligent control system includes a binocular dichoptic module, a covering module, a simulated prism module, a user input module and a human-computer interaction and data management module. The optical display system and the head-mounted strabismus detection system are both connected to the intelligent control system.

2. A full-process intelligent strabismus inspection system as claimed in claim 1, characterized in that: The optical display system comprises a near vision display system and a far vision display system, wherein: The near vision display system includes a near vision display screen, a near vision sight mark light source and a display spectroscope. The near vision display screen and the near vision sight mark are used for adjusting the fixation sight mark and the point light source sight mark respectively. The near vision display screen and the near vision sight mark are imaged near the eyes of the person being measured through the display spectroscope. The telescopic display system includes a telescopic display screen, a telescopic sight mark light source, a display beam splitter and a concave reflector. The telescopic display screen and the telescopic sight mark are used for the adjustable fixation sight mark and the point light source sight mark respectively. The telescopic display screen and the telescopic sight mark are reflected by the display beam splitter into the concave reflector, reflected from the concave reflector, and finally formed through the display beam splitter at a distance in front of the eyes of the person being measured.

3. A full-process intelligent strabismus inspection system as described in claim 1 or 2, characterized in that: The head-mounted strabismus detection system includes a data set annotation module, a pupil detection and positioning module, a line of sight deviation angle calculation module, a head posture detection module and a gaze estimation module, wherein the data set annotation module is used to collect and annotate an eye image data set for training a deep learning model; the pupil detection and positioning module is used to locate the center position of the pupil; the line of sight deviation angle calculation module is used to collect human eye images, and calculate the difference in human eye line of sight angles at two different moments based on the human eye images at two different moments; the head posture detection module is used to detect the head posture of the subject; and the gaze estimation module is used to detect the gaze position of a single eye of the subject on the display screen.

4. A full-process intelligent strabismus inspection system as claimed in claim 3, characterized in that: The implementation process of the dataset annotation module is as follows: Step 111, recruiting data collection subjects, selecting a plurality of persons with normal visual function as data collection subjects; Step 112, data collection, the subject wears a head-mounted strabismus detection device, and the near vision display screen randomly displays an accommodative fixation sight mark. After the subject's line of sight is stable, the eye images of both eyes and the coordinates of the sight mark are saved at this moment; Step 113, prepare the ellipse equation and use the ellipse fitting method to mark the pupil. The standard ellipse equation is ; Step 114, marking pupil edge points, manually marking more than five adjacent points of the pupil and iris; Step 115, construct a matrix equation, substitute each point into the ellipse equation, and obtain a linear equation system; Step 116, solving the problem, using the least square method to solve the overdetermined system of equations, by singular value decomposition or directly calling the numerical calculation library to solve; Step 117, verify the result: draw an ellipse using the calculated parameters A, B, C, D, E, and F, and compare it with the original data points to verify the fitting effect; Step 118, calculating the center point of the ellipse and saving it as the center point of the pupil and the anchor frame; The implementation process of the pupil detection and positioning module is as follows: Step 121, data preparation, using a data set annotation module to prepare a sufficient number of data sets; Step 122, data preprocessing, normalizing and scaling the image for data enhancement to increase the robustness of the model; Step 123, model selection, select a neural network architecture, including U-Net, ResNet or YOLO; Step 124, model training, using the labeled data set to train the model, with the input being the eye image and the output being the pupil anchor box; Step 125, model optimization, adjusting hyperparameters such as network structure, learning rate, batch size, etc. to optimize model recognition accuracy; The implementation process of the sight line deviation angle calculation module is as follows: Step 131, using the pupil detection and positioning module to detect the captured eye image in real time, and saving the images at time t0 and time t1 when the pupil is successfully identified; Step 132, using the pupil detection and positioning module to calculate the center position x0 and x1 of the pupil at time t0 and time t1; Step 133, offset angle , where θ0 is the angle between the measured eye's sight line and the vertical line of the fixation plane at time t0, , R is the radius of the eyeball; The implementation process of the head posture detection module is as follows: Step 141, acquiring data, acquiring real-time (X, Y, Z) angular velocity information from the head posture sensor; Step 142, data processing, integrating the angular velocity data to obtain the angle change of the head, and using a filtering algorithm, such as Kalman filtering or complementary filtering, to reduce the influence of noise; Step 143, posture estimation, using the integrated angle change to estimate the posture of the head, namely pitch, roll and yaw, and performing posture correction in combination with the accelerometer data; Step 144, implement posture detection, set a threshold or pattern recognition algorithm according to the posture change, and detect specific head movements.

5. A full-process intelligent strabismus inspection system as claimed in claim 1 or 2, characterized in that: The intelligent control system includes a binocular dichoptic module, a covering module, a simulated prism module, a user input module and a human-computer interaction and data management module, wherein the binocular dichoptic module is used to separate the images observed by the person being tested, so that when the person being tested observes the same display plane, both eyes can obtain different images; the covering module is used to achieve the shielding of the observing eyes of the person being tested; the simulated prism module is used to replace the traditional prism to achieve the function of accurately offsetting the image observed by the strabismus eyes of the person being tested, and to achieve accurate measurement of the strabismus of the person being tested; the user input module is used to assist the person being tested in moving the visual mark in the image, and to achieve the measurement of the subjective strabismus; the human-computer interaction and data management module is used to assist the tester in using the equipment, and to process, store and print the data after the measurement is completed.

6. A full-process intelligent strabismus inspection system as claimed in claim 5, characterized in that: The covering module is composed of a liquid crystal light valve module and a binocular vision module. The implementation process of the covering module is as follows: Step 211, selecting the occluded eye; Step 212, closing the display image of the blocked eye in the binocular dichoptic module; Step 213, closing the liquid crystal light valve module that blocks the eyes.

7. A full-process intelligent strabismus inspection system as claimed in claim 6, characterized in that: The liquid crystal light valve module includes a polarizing layer, a liquid crystal layer and a polarizing layer. The blocking of the liquid crystal light valve module will not affect the camera to capture the eye image. The implementation process of the liquid crystal light valve module is as follows: Step 221, keeping the polarization directions of the polarizing layer and the analyzing layer consistent; Step 222, when no electric field is applied to the liquid crystal layer, the light of the display entering from the polarizing layer does not change its polarization direction after passing through the liquid crystal layer, and can pass through the analyzer layer; Step 223, when the controller applies an electric field to the liquid crystal layer, the light entering from the polarizing layer has a polarization direction perpendicular to the original direction after passing through the liquid crystal layer, and cannot pass through the analyzing layer, thereby achieving a shielding effect; Step 224, after the light from the eye of the person under test passes through the polarization analyzer layer and the liquid crystal layer, it is directly reflected into the camera by the camera beam splitter, so the camera can capture the image regardless of whether the liquid crystal layer applies an electric field; The implementation process of the simulated prism module is as follows: Step 231: selecting a near vision display screen or a far vision display screen as a display screen for simulating a prism module, and the distance between the virtual image of the display screen after passing through the optical path of the optical display system and the eyes of the person being measured is d; Step 232: using a binocular vision module to separate the images observed by the left and right eyes of the person under test, so that the left eye of the person under test can only observe the left eye observation image of the screen, and the right eye can only observe the right eye observation image of the screen; Step 233: By shifting the position of the strabismus eye observing the screen, the observation effect of the person under test using a prism is simulated. Assuming that the degree of the prism to be placed is △, the separation of the sight mark is L, and the angle between the main eye's line of sight and the perpendicular line of the display plane is θ, then , where L is in cm and d is in m.

8. A method for implementing the full-process intelligent strabismus inspection system as claimed in claim 1, characterized in that: The method comprises an eye movement function examination process and a binocular vision function examination process; The eye movement function test process includes the cover-uncover method, the alternating cover method, the prism cover method, the simulated corneal light reflection method and the nine eye position test method; The binocular vision function test process includes Worth four-point light test method, Bagolini linear mirror test method and stereoscopic test method.

9. The method according to claim 8, characterized in that During the eye movement function test, the cover-uncover method is used to measure the strabismus of a single eye. The implementation process of the cover-uncover method is as follows: Step 311, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 312, displaying an accommodative fixation target on the near vision display screen, and recording the image I_0 of the eye being tested at this moment after the line of sight of the person being tested is stabilized; Step 313, using a covering module to cover the non-tested eye on the opposite side of the tested eye, and recording the image I_1 of the tested eye at this moment after the tested person's line of sight is stabilized; Step 314, using a sight deviation angle calculation module to calculate the gaze deviation angle of the measured eye from image I_0 to image I_1; Step 315, repeating the measurement process from step 312 to step 313 for multiple times, and calculating the average of the multiple results, to obtain the average gaze deviation angle of the measured eye as the myopia degree of the measured eye; Step 316, if the measured eye does not move or the strabismus is too small, measure the other eye in the same way; Step 317, changing the near vision display screen in step 312 to a far vision display screen or a far vision sight mark light source, repeating steps 312 to 316, and obtaining the far vision strabismus of the eye being tested; The implementation process of the alternating covering method is as follows: Step 321, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 322, displaying an accommodative fixation sight mark on the near vision display screen, and waiting for the line of sight of the person being measured to stabilize; Step 323, using a covering module to cover the left eye, waiting for the person being tested to stabilize his / her sight line, and then recording the left eye image I_left_0 and the right eye image I_right_0 at this moment; Step 324, using the covering module to cover the right eye, waiting for the measured person's sight to stabilize, and then recording the left eye image I_left _1 and the right eye image I_right _1 at this moment; Step 325, using the sight deviation angle calculation module to calculate the gaze deviation angle of the measured eye from image I_left_1 to image I_left_0, recorded as the strabismus of the left eye this time; Step 326, using the sight deviation angle calculation module to calculate the gaze deviation angle of the eye under test from image I_right_0 to image I_right_1, recorded as the strabismus of the right eye this time; Step 327, repeating the measurement process from step 322 to step 326 for multiple times, and calculating the average of the multiple results, to obtain the average gaze deviation angle of the left and right eyes as the strabismus of the measured eye; Step 328, changing the near vision display screen in step 322 to a far vision display screen or a far vision sight mark light source, repeating steps 322 to 327, and obtaining the far vision strabismus of the left and right eyes; The implementation process of the prism covering method is as follows: Step 331, initialize array D i Used to store the measured strabismus, where i is the number of times the cover-uncover method or the alternating cover method is used after adding the prism, i = 0, 1, 2, 3 ...; set D0 to the strabismus measured without adding the prism or the strabismus input by the user, and set the prism degree D at this moment to D0; Step 332, implement the cover-uncover method or the alternating cover method to measure the strabismus degree D of the person under test under the prism. i ; Step 333, if the strabismus degree D is measured i is 0, then the strabismus of the person being measured at this moment is If the measured strabismus degree D i If it is not 0, then set the prism degree D at this moment to Then repeat step 332; The implementation process of the simulated corneal light reflection method is as follows: Step 341, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 342, displaying an accommodative fixation sight mark on the near vision display screen, and waiting for the line of sight of the person being measured to stabilize; Step 343, using a gaze estimation module to calculate the gaze positions of the left and right eyes on the screen; Step 344, calculate the degree of strabismus of the left and right eyes according to the deviation between the left and right eyes' gaze positions on the screen and the actual sight mark. Assuming that the distance from the sight mark to the eyes of the person being tested is d, and the distance from the gaze point to the sight mark is L, then the strabismus degree D is m =L / d, where L is in cm and d is in m; Step 345, changing the near vision display screen in step 342 to a far vision display screen or a far vision sight mark light source, repeating steps 342 to 344, and obtaining the far vision strabismus of the left and right eyes; The implementation process of the nine-eye position inspection method is as follows: Step 351, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement begins, a voice announcer is used to guide the person being measured to look at the sight mark that appears in front of the eyes during the measurement process; Step 352, setting nine eye position sight mark positions on the near vision display screen, respectively in the original position where the person being measured is looking, horizontally to the right, horizontally to the left, to the upper left, to the lower left, to the upper right, to the lower right, vertically upward, and vertically downward; Step 353, lighting up the sight mark accordingly, and implementing the cover-uncover method, the alternating cover method, the prism cover method or the simulated corneal light reflection method to measure the myopia degree of the person being measured at the nine eye positions; Step 354, setting a sight mark at the center of the telescopic display screen; Step 355, guide the person being measured to rotate his head in the original position, horizontally to the right, horizontally to the left, to the upper left, to the lower left, to the upper right, to the lower right, vertically upward, and vertically downward directions respectively through the voice announcer. After the head posture detection module confirms that the head posture is correct, implement the cover-uncover method, the alternating cover method, the prism cover method, or the simulated corneal light reflection method to measure the degree of hypertropia of the person being measured in the nine eye positions.

10. The method according to claim 8, characterized in that During the binocular vision function test, the implementation process of the Worth four-light test method is as follows: Step 411, after the person being measured wears the head-mounted strabismus detection device, the person being measured places his head on a head fixing bracket, and before the measurement starts, the person being measured is guided by a voice announcer to observe the image displayed on the display screen; Step 412, lighting the Worth four-point light in the near-sighted sight mark light source, wherein the right-eye image displays sight mark (1) and sight mark (3), and the left-eye image displays sight mark (2), sight mark (3), and sight mark (4); Step 413, allowing the person being tested to select the image that he or she observes, and observing different images corresponds to different visual function problems of the person being tested; Step 414, changing the near vision sight mark light source in step 412 to the far vision sight mark light source, repeating steps 412 and 413 to obtain the far vision measurement result; The implementation process of the Bagolini linear microscopy method is as follows: Step 421, the person being measured wears the head-mounted strabismus detection device, turns off the binocular dichoptic module, installs the Bagolini linear mirror on the replaceable lens holder, places the head on the head fixing bracket, and guides the person being measured to observe the image displayed on the display screen through the voice announcer before the measurement starts; Step 422, lighting a white light source for Bagolini linear speculum examination in the nearsightedness target light source, judging the visual acuity of the person being tested according to the person being tested's feedback, a person with normal binocular single vision function sees a complete cross, a patient with monocular gaze can see a cross when binocularly gazing due to peripheral fusion, but because the macula of the non-dominant eye is suppressed, the center of the stripe seen by the non-dominant eye is discontinuous; Step 423, changing the nearsightedness sight mark light source in step 422 to a farsightedness sight mark, repeating step 422, and obtaining the inspection result under the farsightedness condition; The implementation process of the stereoscopic inspection method is as follows: Step 431, after the person being measured wears the head-mounted strabismus detection device, the person's head is placed on a head fixing bracket, and before the measurement starts, the person being measured is guided by a voice announcer to observe the image displayed on the display screen; Step 432, combining with the binocular dichoptic module, displaying near stereoscopic vision inspection images such as Titmus stereogram and TNO stereogram on the near vision display screen to inspect the near stereoscopic vision function of the person being tested; Step 433, in combination with the binocular dichoptic module, a far stereoscopic vision test chart such as a Randot test chart and a Frisby-Davis test chart is displayed on the far vision display screen to check the far stereoscopic vision function of the person being tested.

Images