Eye chamber angle state determination method and device, equipment and storage medium
By acquiring the eye image and calculating its area ratio, determining the atrial angle state of the eye, the problem of inaccurate results in the prior art is solved, and the accuracy and objectivity of the results are improved.
Patent Information
- Application Number
- CN202310983875.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-07
- Publication Date
- 2025-02-14
AI Technical Summary
In the prior art, the determination result of the ocular atrial state is inaccurate, which affects the accuracy of the result.
By acquiring the eye image of the target object, the area of the first area and the second area in the image is determined, and the ocular atrial angle state is determined based on the ratio of the area of the two. This method eliminates individual differences and improves the accuracy of results.
It improves the accuracy of the determination results of the angle of the eye, reduces subjectivity, and enhances objectivity.
Smart Images
Figure CN119423681A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image processing, and in particular to a method, device, equipment and storage medium for determining the state of an eye chamber angle. Background Art
[0002] In people's eyes, the angle structure of the anterior chamber of the eye refers to the area formed by the iris and the corneal limbus, which is the way for the aqueous humor to flow out. The state of the eye angle of people usually includes open angle and closed angle.
[0003] In the prior art, professionals usually judge the state of people's eye angles based on their own manual experience. However, in the process of manually determining the state of the eye angle, the central anterior chamber depth (ACD) is usually used as an indicator to determine the state of the eye angle. The central anterior chamber depth is not clearly represented in the structure of the angle of the anterior chamber of the eye of people, and the determination result of the eye angle state of people may be inaccurate, which may easily affect the accuracy of the result of the eye angle state.
[0004] Therefore, how to improve the accuracy of the results of determining the state of people's eye chamber angles has become a problem that needs to be solved urgently. Summary of the invention
[0005] In view of this, the embodiments of the present application provide a method, apparatus, device and storage medium for determining the state of the eye angle, aiming to improve the accuracy of the determination result of the eye angle state.
[0006] In a first aspect, an embodiment of the present application provides a method for determining an eye chamber angle state, the method comprising:
[0007] Acquire an eye image of a target object; the eye image includes a first light ray and a second light ray; the first light ray is a light ray formed by irradiating the cornea of the target object; the second light ray is a light ray formed by projecting the first light ray through the cornea of the target object onto the iris of the target object and passing through the pupil of the target object;
[0008] Determine the area of a first region in the eye image, and determine the area of a second region in the eye image; the first region in the eye image is a closed region formed by the first light and the second light on the eye image; the second region in the eye image is a cornea region in the eye image;
[0009] The state of the eye angle of the target object is determined based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image.
[0010] Optionally, the determining the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image specifically includes:
[0011] Based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image and a target classifier, the eye angle state of the target object is determined; the threshold of the target classifier is determined by a receiver operating characteristic curve; the receiver operating characteristic curve is drawn based on multiple eye image samples.
[0012] Optionally, the determining the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image and a target classifier specifically includes:
[0013] Determine a ratio of an area of a first region in the eye image to an area of a second region in the eye image to obtain an area ratio;
[0014] The area ratio is input into the target classifier to obtain the eye chamber angle state of the target object.
[0015] Optionally, the eye image includes a first eye image, a second eye image, a third eye image and a fourth eye image corresponding to the same side eye of the target object;
[0016] wherein the first light in the first eye image is above the pupil in the first eye image;
[0017] The first light in the second eye image is below the pupil in the second eye image;
[0018] The first light in the third ocular image is to the left of the pupil in the third ocular image;
[0019] The first light in the fourth ocular image is to the right of the pupil in the fourth ocular image;
[0020] The determining the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image and a target classifier specifically includes:
[0021] Determine a ratio of an area of a first region to an area of a second region in the first eye image, the second eye image, the third eye image, and the fourth eye image, respectively, to obtain a first area ratio corresponding to the first eye image, a second area ratio corresponding to the second eye image, a third area ratio corresponding to the third eye image, and a fourth area ratio corresponding to the fourth eye image;
[0022] Averaging the sum of the first area ratio, the second area ratio, the third area ratio and the fourth area ratio to obtain a ratio average;
[0023] The ratio average is input into the target classifier to obtain the eye chamber angle state of the target object.
[0024] Optionally, it is characterized in that the receiver operating characteristic curve is drawn in the following manner, specifically including:
[0025] Acquire the plurality of eye image samples; the plurality of eye image samples include positive eye image samples and negative eye image samples; the positive eye image samples are eye images that have been determined to have an open angle; the negative eye image samples are eye images that have been determined to have a closed angle or a suspected closed angle;
[0026] In each eye image sample, determining a ratio of an area of the first region to an area of the second region, and obtaining an area ratio corresponding to each eye image sample;
[0027] Inputting the area ratio corresponding to each eye image sample into a classifier to obtain a prediction probability corresponding to each eye image;
[0028] The receiver operating characteristic curve is drawn according to the prediction probability corresponding to each eye image and different preset thresholds.
[0029] Optionally, the determining the area of the first region in the eye image and determining the area of the second region in the eye image specifically includes:
[0030] Converting the eye image into an eye grayscale image;
[0031] Binarizing the eye grayscale image to obtain a plurality of edge pixel points in the eye grayscale image;
[0032] Connecting edge pixel points on the first light and edge pixel points on the second light to obtain a first region in the eye image; and connecting edge pixel points on the edge of the cornea to obtain the second region;
[0033] The number of pixels in the first region of the eye image is determined as the area of the first region of the eye image; and the number of pixels in the second region of the eye image is determined as the area of the second region of the eye image.
[0034] In a second aspect, an embodiment of the present application provides a device for determining an eye chamber angle state, the device comprising:
[0035] An image acquisition module, used for acquiring an eye image of a target object; the eye image includes a first light and a second light; the first light is a light formed by irradiating the cornea of the target object; the second light is a light formed by the first light passing through the cornea of the target object, projecting onto the iris of the target object, and passing through the pupil of the target object;
[0036] an area determination module, used to determine the area of a first area in the eye image and to determine the area of a second area in the eye image; the first area in the eye image is a closed area formed by the first light and the second light on the eye image; the second area in the eye image is a cornea area in the eye image;
[0037] The eye angle state determination module is used to determine the eye angle state of the target object based on the ratio of the area of the first area in the eye image to the area of the second area in the eye image.
[0038] Optionally, the eye chamber angle state determination module is specifically used to:
[0039] Based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image and a target classifier, the eye angle state of the target object is determined; the threshold of the target classifier is determined by a receiver operating characteristic curve; the receiver operating characteristic curve is drawn based on multiple eye image samples.
[0040] In a third aspect, an embodiment of the present application provides a device for determining an eye chamber angle state, the device comprising a memory and a processor:
[0041] The memory is used to store a computer program and transmit the computer program to the processor;
[0042] The processor is used to execute the computer program so that the device executes the method for determining the state of the eye chamber angle as described in the first aspect.
[0043] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed, the device executing the computer program implements the method for determining the state of the eye chamber angle as described in the first aspect above.
[0044] Compared with the prior art, the embodiments of the present application have the following beneficial effects:
[0045] The embodiments of the present application provide a method, apparatus, device and storage medium for determining the state of an eye angle. In the method, an eye image of a target object is obtained; the eye image includes a first light ray and a second light ray; the first light ray is a light ray formed by irradiating the cornea of the target object; the second light ray is a light ray formed by the first light passing through the cornea of the target object and projected onto the iris of the target object, and passing through the pupil of the target object; the area of a first area in the eye image is determined, and the area of a second area in the eye image is determined; the first area in the eye image is a closed area formed by the first light ray and the second light ray on the eye image; the second area in the eye image is a corneal area in the eye image; based on the ratio of the area of the first area in the eye image to the area of the second area in the eye image, the state of the eye angle of the target object is determined.
[0046] It can be seen that using the area of the first region (i.e., the anterior chamber cross-sectional area) as an indicator for determining the state of the eye angle of the target object improves the characterization of the eye angle state of the target object, and the ratio of the area of the first region to the area of the second region can eliminate individual differences and obtain a more accurate determination result of the eye angle state of the target object, thereby improving the accuracy of the determination result of the eye angle state.
[0047] In addition, the present application does not require manual determination of the eye angle state of the target object, which can avoid the problem of excessive subjectivity in the process of determining the eye angle state, improve the objectivity of the process of determining the eye angle state, and thus further improve the accuracy of the determination result of the eye angle state. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the technical solutions in this embodiment or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0049] Figure 1 An application scenario of a method for determining an eye chamber angle state provided in an embodiment of the present application;
[0050] Figure 2 A flow chart of a method for determining an eye chamber angle state provided in an embodiment of the present application;
[0051] Figure 3 A schematic diagram of an eye image provided in an embodiment of the present application;
[0052] Figure 4 A schematic diagram of an image sample of a target area positioning model provided in an embodiment of the present application;
[0053] Figure 5 A schematic diagram of an eye structure provided in an embodiment of the present application;
[0054] Figure 6a A schematic diagram of a first eye image provided in an embodiment of the present application;
[0055] Figure 6b A schematic diagram of a second eye image provided in an embodiment of the present application;
[0056] Figure 6c A schematic diagram of a third eye image provided in an embodiment of the present application;
[0057] Figure 6d A schematic diagram of a fourth eye image provided in an embodiment of the present application;
[0058] Figure 7 A schematic diagram of the structure of a device for determining the state of an eye chamber angle provided in an embodiment of the present application. DETAILED DESCRIPTION
[0059] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0060] First, several terms that may be involved in the embodiments of the present application are explained below.
[0061] The anterior chamber area (ACA) refers to the cross-sectional area enclosed by the corneal endothelium, the anterior surface of the iris, and the anterior surface of the lens located within the pupil.
[0062] Receiver operating characteristic curve (ROC curve for short) refers to the line connecting the points under specific stimulation conditions, with the false alarm probability P(y / N) obtained by the subjects under different judgment criteria as the horizontal coordinate and the hit probability P(y / SN) as the vertical coordinate.
[0063] Area Under Curve (AUC): It is defined as the area under the ROC curve and the coordinate axis. Since the horizontal and vertical axes of the ROC curve have a value range of [0, 1], the AUC is ≤ 1. AUC is an indicator to measure the pros and cons of different evaluation criteria in the above ROC curves. Different evaluation criteria correspond to different AUCs. The larger the AUC, the higher the authenticity of the corresponding evaluation criteria.
[0064] The state of people's eye angles is usually divided into open angle and closed angle. At present, manual methods are usually used to determine the state of people's eye angles. Generally speaking, professionals can use tools such as slit lamps to manually observe the central anterior chamber depth of people's eyes, and then manually judge whether people's eye angles are normal based on the van Herick grading method. However, this method uses the central anterior chamber depth as an indicator to determine the state of the eye angle. The central anterior chamber depth is not clearly represented in the angle structure of the anterior chamber of the eye of people, which easily affects the accuracy of the results of the eye angle state.
[0065] Based on this, in order to solve the above problems, the embodiments of the present application provide a method, device, equipment and storage medium for determining the state of the eye angle, wherein the method obtains an eye image of the target object; determines the area of the first area in the eye image, and determines the area of the second area in the eye image; and determines the state of the eye angle of the target object based on the ratio of the area of the first area in the eye image to the area of the second area in the eye image. It can be seen that the area of the first area (i.e., the anterior chamber cross-sectional area) is used as an indicator for determining the state of the eye angle of the target object. This indicator can improve the characterization of the state of the eye angle of the target object. Based on the ratio of the area of the first area to the area of the second area, individual differences can be eliminated, and a more accurate determination result of the state of the eye angle of the target object can be obtained, thereby improving the accuracy of the determination result of the state of the eye angle.
[0066] In addition, the present application does not require manual determination of the target object's eye angle state, which can avoid the problem of excessive subjectivity in the process of determining the eye angle state, improve the objectivity of the process of determining the eye angle state, and further improve the accuracy of the determination result of the eye angle state.
[0067] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the following first describes the application scenario of the technical solution of the present application. One of the scenarios of the embodiment of the present application can be applied to Figure 1The scene shown in the figure includes a mobile phone slit lamp 110 and a server 120, wherein the mobile phone slit lamp 110 includes a light source for providing a first light beam irradiated onto the cornea of the target object, and the mobile phone slit lamp 110 also includes a camera for capturing an eye image of the target object, and the server 120 uses the implementation method provided in the embodiment of the present application to obtain the eye image of the target object captured by the mobile phone slit lamp 110, and uses the implementation method provided in the embodiment of the present application to determine the state of the eye angle of the target object.
[0068] First, in the above application scenario, although the actions of the implementation provided by the embodiment of the present application are described as being executed by the server 120; however, the embodiment of the present application is not limited in terms of the execution subject, as long as the actions disclosed in the implementation provided by the embodiment of the present application are executed. For example, the method for determining the state of the eye chamber angle provided by the embodiment of the present application can be executed by a terminal device with data processing capabilities; it can also be executed collaboratively by the terminal device and the server.
[0069] Secondly, the above scenario is only an example scenario provided by the embodiment of the present application, and the embodiment of the present application is not limited to this scenario.
[0070] The specific implementation of the method, device, equipment and storage medium for determining the eye chamber angle state in the embodiments of the present application will be described in detail below with reference to the accompanying drawings by way of examples.
[0071] See also Figure 2 , which is a flow chart of a method for determining an eye chamber angle state provided in an embodiment of the present application, combined with Figure 2 As shown, it may specifically include:
[0072] S201: Acquire an eye image of a target object.
[0073] The eye image includes a first light and a second light; the first light is the light formed by irradiating the cornea of the target object; the second light is the light formed by the first light passing through the cornea of the target object and projected onto the iris of the target object, and passing through the pupil of the target object. The target object refers to any person with a certain eye angle state, and this application does not limit this.
[0074] It should be noted that the eye image includes the anterior segment structure of the target object, and therefore may also be referred to as an anterior segment image.
[0075] As an example, the light (i.e., the first light) emitted by the light source of the mobile phone slit lamp in the above application scenario can be irradiated onto the cornea of the target object. The cornea is nearly transparent, so the first light can pass through the cornea of the target object and be mapped onto the iris. Then the mobile phone slit lamp can be moved so that the second light moves with the movement of the first light until the second light passes through the pupil of the target object. Then the eye of the target object is photographed to obtain the eye image, so that the server or terminal device of the embodiment of the present application can obtain the eye image.
[0076] S202: Determine the area of the first region in the eye image, and determine the area of the second region in the eye image.
[0077] The first area in the eye image is a closed area formed by the first light and the second light on the eye image; the second area in the eye image is a cornea area in the eye image.
[0078] For ease of understanding, the following Figure 3 The schematic diagram of the eye image shown in FIG. 1 specifically describes the first area and the second area in the eye image. Figure 3 As shown, the area enclosed by the first light and the second light is the first area, and the cornea area is the second area.
[0079] It is understandable that, generally speaking, people's upper and lower eyelids will cover some actual corneal areas, but when determining the area of the corneal area in the eye image (the second area), the area of the area covered by the eyelids can be inferred by image processing algorithms such as image de-occlusion, and thus the area of the complete second area can be determined.
[0080] In a possible implementation of the present application, the process of a terminal device or a server determining a first area and a second area in an eye image may specifically include: first converting the eye image into an eye grayscale image; then binarizing the eye grayscale image to obtain a plurality of edge pixel points in the eye grayscale image; the edge pixel points located on the first light ray and the edge pixel points located on the second light ray are connected to form a first area, and the edge pixel points located at the edge of the cornea are connected to form a second area.
[0081] As an example, the eye image can be read by the imread function in the opencv-python library and converted into an eye grayscale image. After black and white binarization processing, edge pixels are obtained to determine the first area and the second area in the image.
[0082] In order to accurately determine the first region and the second region in the eye image, in a possible implementation of the present application, a target region positioning model may be trained to position the first region and the second region in the eye image of the target object.
[0083] As an example, a technician may use a labeling tool to label the cornea region (second region) in the eye image, and the first light and the second light in the eye image, as image samples, to train the target region positioning model to be trained, so that the first region and the second region in the eye image can be located by the target region positioning model. For example, the labeling tool may be a Labelme labeling tool. Figure 4 , which is a schematic diagram of an image sample of a target area positioning model provided in an embodiment of the present application.
[0084] In a possible implementation of the present application, the area of the first region can be represented by the number of pixels in the first region in the eye image. Similarly, the area of the second region can also be represented by the number of pixels in the second region in the eye image, which is conducive to the execution of the subsequent step of calculating the ratio of the area of the first region to the area of the second region.
[0085] S203: Determine the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image.
[0086] In this application, the anterior chamber cross-sectional area ACA is used as an important indicator to characterize the state of the eye's chamber angle. Figure 5 The schematic diagram of the eye structure shown in FIG. Figure 5 As shown, the cross-sectional area of the area surrounded by the corneal endothelium and the anterior surface of the iris and the anterior surface of the pupil area lens is the anterior chamber cross-sectional area. The area of the first area is the anterior chamber cross-sectional area. It should be noted that in the embodiment of the present application, the eye angle state is the angle state of the anterior chamber of the eyes of people.
[0087] In a possible implementation of the present application, S203 may specifically be: determining the eye angle state of the target object based on the ratio of the area of the first area in the eye image to the area of the second area in the eye image and a target classifier; the threshold of the target classifier is determined by a receiver operating characteristic curve; and the receiver operating characteristic curve is drawn based on multiple eye image samples.
[0088] It can be seen that using the threshold of the target classifier as a unified standard can avoid the problem of excessive subjectivity in the process of determining the eye angle state, and improve the objectivity in the process of determining the eye angle state. At the same time, using the threshold determined by the receiver operating characteristic curve as the threshold of the target classifier, and the receiver operating characteristic curve is drawn based on multiple eye image samples, can improve the classification accuracy of the target classifier, thereby improving the accuracy of the results of the eye angle state.
[0089] In practical applications, considering that the size of the acquired eye image is uncertain, there may be errors in the measurement of the anterior chamber cross-sectional area due to the difference in the size of the eye image. Therefore, the embodiment of the present application uses the ratio of the area of the first area to the area of the second area, that is, the area ratio, to characterize the state of the eye chamber angle.
[0090] In one possible method of the present application, a receiver operating characteristic curve can be drawn through the following steps 1 to 4 to determine the threshold.
[0091] Step 1: Get multiple eye image samples.
[0092] Among them, the multiple eye image samples include positive eye image samples and negative eye image samples; the positive eye image samples are eye images that have been determined to have an open angle; the negative eye image samples are eye images that have been determined to have a closed angle or a suspected closed angle.
[0093] As an example, eye images of both eyes can be collected from 500 people whose eye angle status is open (normal angle status), and 1000 eye images can be obtained as positive eye image samples; similarly, eye images of both eyes can be collected from 500 people whose eye angle status is suspected closed angle or closed angle (abnormal angle status), and 1000 eye images can be obtained as negative eye image samples.
[0094] Among them, in order to improve the accuracy of the eye angle state corresponding to the eye image samples, the gold standard gonioscope is used as the reference standard to identify the angle closure. The eyes of the personnel are examined in a dark room and under static gonioscope. If the functional trabecular structure is not visible ≥180°, it means the angle is suspected to be closed or the angle is closed. Otherwise, it means the angle is open.
[0095] Step 2: In each eye image sample, determine the ratio of the area of the first region to the area of the second region to obtain an area ratio corresponding to each eye image sample;
[0096] In a possible implementation of the present application, based on the above example, for each eye image sample, the number of pixels in the first region can be calculated separately, and then the number of pixels in the second region can be calculated, and the ratio of the number of pixels in the first region to the number of pixels in the second region can be determined as the ratio of the area of the first region to the area of the second region, and the area ratio corresponding to each eye image sample can be obtained.
[0097] Step 3: Input the area ratio corresponding to each eye image sample into the classifier to obtain the prediction probability corresponding to each eye image.
[0098] The classifier can predict the state of the chamber angle based on the area ratio of the image sample. As an implementation, the eye image sample is input into the classifier, and the output result of the classifier includes a probability value from 0 to 1. For example, the classifier can be at least one of a decision tree classifier, a support vector machine classifier, a naive Bayes classifier, or a K-nearest neighbor classifier, and the present application does not limit this.
[0099] Step 4: Draw the receiver operating characteristic curve based on the prediction probability corresponding to each eye image and different preset thresholds.
[0100] The chamber angle state of the eye image sample with a probability value greater than the threshold is predicted as an open chamber angle, and the chamber angle state of the eye image sample with a probability value less than or equal to the threshold is predicted as a suspected closed chamber angle or a closed chamber angle. This application does not limit this.
[0101] Therefore, by inputting the area ratio corresponding to each eye image sample into the classifier, the chamber angle states of multiple eye image samples can be classified for the second time. At this time, each eye image sample will have an actual chamber angle state (a positive eye image sample or a negative eye image sample) and a predicted chamber angle state (based on the prediction of the classifier).
[0102] It should be noted that different preset thresholds may be multiple thresholds of multiple classifiers, or different thresholds of a classifier, and this application does not limit this.
[0103] As an example, a classifier has two different preset thresholds, A1 and A2. Eye image sample 1 is a negative eye image sample, and its corresponding classifier output result is a1 (the probability value mentioned above). Eye image sample 2 is a positive eye image sample, and its corresponding classifier output result is a2. There is a numerical relationship of A1<a1<a2<A2. Then the actual chamber angle state of eye image 1 is closed, and the predicted chamber angle state obtained by comparison with A1 is open, and the predicted chamber angle state obtained by comparison with A2 is closed or suspected closed; the actual chamber angle state of eye image 2 is open, and the predicted chamber angle state obtained by comparison with A1 is open, and the predicted chamber angle state obtained by comparison with A2 is closed or suspected closed.
[0104] Among them, the false positive samples of eye images in the eye image samples are when the actual chamber angle state is suspected chamber angle closure or chamber angle closure, but the predicted chamber angle state is open; the true positive samples of eye images in the eye image samples are when the actual chamber angle state is open chamber angle, but the predicted chamber angle state is open.
[0105] Then, according to the steps of drawing the receiver operating characteristic curve, for each preset threshold, the false positive rate (False Positive Rate, FPR) in multiple image samples is calculated, that is, the ratio of false positive samples of eye images to negative samples of eye images; then the true positive rate (TruePositiveRate, TPR) in multiple image samples is calculated, that is, the ratio of true samples of eye images to positive samples of eye images. Each preset threshold will correspond to a (FPR, TPR). Using FPR as the horizontal axis and TPR as the vertical axis, draw the (FPR, TPR) corresponding to each preset threshold, and connect all (FPR, TPR) to obtain the receiver operating characteristic curve.
[0106] In the receiver operating characteristic curve, a preset threshold value that maximizes the area AUC enclosed by the coordinate axis under the ROC curve is used as the threshold value of the target classifier to determine the chamber angle state of the eye image of the target object.
[0107] In the embodiment of the present application, there are multiple possible implementations of the above-mentioned S203, which are introduced below. It should be noted that the implementations given in the following introduction are only for illustrative purposes and do not represent all implementations of the embodiment of the present application.
[0108] Then a first optional implementation method of S203 is: determine the ratio of the area of the first region in the eye image to the area of the second region in the eye image to obtain an area ratio; input the area ratio to a target classifier to obtain the eye angle state of the target object.
[0109] A second optional implementation of S203 is: the eye image includes a first eye image, a second eye image, a third eye image, and a fourth eye image corresponding to the same side eye of the target object; wherein the first light in the first eye image is above the pupil in the first eye image; the first light in the second eye image is below the pupil in the second eye image; the first light in the third eye image is to the left of the pupil in the third eye image; and the first light in the fourth eye image is to the right of the pupil in the fourth eye image;
[0110] Then S203 may specifically include: determining the ratio of the area of the first region to the area of the second region in the first eye image, the second eye image, the third eye image, and the fourth eye image, respectively, to obtain a first area ratio corresponding to the first eye image, a second area ratio corresponding to the second eye image, a third area ratio corresponding to the third eye image, and a fourth area ratio corresponding to the fourth eye image; averaging the sum of the first area ratio, the second area ratio, the third area ratio, and the fourth area ratio to obtain a ratio average; inputting the ratio average into the target classifier to obtain the eye angle state of the target object. It can be seen that by comprehensively considering the area ratios of different directions of the same side of the eye of the target object, the error caused by different results of the eye angle state obtained by the area ratios corresponding to different methods of the target object can be avoided, and a more accurate eye angle state can be obtained.
[0111] It should be noted that the eye image can also be at least one of the first eye image, the second eye image, the third eye image and the fourth eye image, and the present application does not impose any limitation on this.
[0112] As an example, see Figure 6a-6d , respectively represent schematic diagrams of the first eye image, the second eye image, the third eye image, and the fourth eye image in the embodiment of the present application. Figure 6a As shown, Figure 6a is the first eye image, Figure 6a The first ray in the first eye image is above the pupil; Figure 6b As shown, Figure 6b is the second eye image, Figure 6b The first ray in the image is below the pupil in the second eye image; Figure 6c As shown, Figure 6c is the third eye image, Figure 6c The first ray in is to the left of the pupil in the third eye image; Figure 6d As shown, Figure 6d is the fourth eye image, Figure 6d The first ray in is to the right of the pupil in the fourth eye image.
[0113] It is understandable that when drawing the receiver operating characteristic curve, it is also possible to take images of four directions such as the first eye image, the second eye image, the third eye image and the fourth eye image for each person's eye, and calculate the ratio average of the four eye image samples on the same side of the eye of the same person, and input them into the classifier to obtain the predicted eye angle state. The specific steps can be referred to the above example, and will not be repeated here. The receiver operating characteristic curve thus drawn can determine a more appropriate threshold to further improve the accuracy of the eye angle state of the eye image.
[0114] In the related art, professionals usually use tools such as slit lamps to manually observe the ACD of people's eyes, and use the central anterior chamber depth as an indicator to determine the state of people's eye angles. In this application, ACA is used as an indicator to determine the state of the eye angle. Compared with ACD, ACA is more clearly represented in the eye image, and the ratio of the area of the first area to the area of the second area can eliminate individual differences between different target objects, so as to obtain a more accurate determination result of the eye angle state of the target object, which is conducive to further improving the accuracy of the determination result of the eye angle state.
[0115] In addition, the related art relies on experienced professionals, and when professionals manually determine the state of the eye's chamber angle, there is a lack of unified determination standards. There is a strong subjectivity in the process of determining the state of the eye's chamber angle, which affects the objectivity of the results of the eye's chamber angle state. In this application, the threshold of the target classifier is used as a unified standard to avoid the problem of excessive subjectivity in the process of determining the state of the eye's chamber angle, thereby improving the objectivity of the process of determining the state of the eye's chamber angle. At the same time, the threshold determined by the receiver operating characteristic curve is used as the threshold of the target classifier, and the receiver operating characteristic curve is drawn based on multiple eye image samples, which can improve the classification accuracy of the target classifier, thereby improving the accuracy of the results of the eye's chamber angle state.
[0116] In addition, the eye image in the embodiment of the present application can be obtained by taking pictures with a mobile phone slit lamp, and is no longer limited to professional detection equipment to detect the state of the eye angle. It can be applied to multiple application scenarios such as daily community detection and remote detection. The eye image taken by the mobile phone slit lamp can be determined by the method for determining the state of the eye angle provided in the embodiment of the present application, which brings people a convenient and fast experience.
[0117] The above are some specific implementations of the method for determining the state of the eye chamber angle provided in the embodiments of the present application. Based on this, the present application also provides a corresponding device. The device provided in the embodiments of the present application will be introduced from the perspective of functional modularization.
[0118] See also Figure 7, which is a schematic diagram of the structure of a device 700 for determining an eye chamber angle state provided in an embodiment of the present application, and the device 700 may include:
[0119] The image acquisition module 710 is used to acquire an eye image of the target object; the eye image includes a first light and a second light; the first light is a light formed by irradiating the cornea of the target object; the second light is a light formed by the first light passing through the cornea of the target object, projecting onto the iris of the target object, and passing through the pupil of the target object;
[0120] The area determination module 720 is used to determine the area of a first area in the eye image and the area of a second area in the eye image; the first area in the eye image is a closed area formed by the first light and the second light on the eye image; the second area in the eye image is a cornea area in the eye image;
[0121] The eye angle state determination module 730 is used to determine the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image.
[0122] As an implementation manner, the eye chamber angle state determination module can be specifically used for:
[0123] Based on the ratio of the area of a first region in the eye image to the area of a second region in the eye image and a target classifier, the state of the eye angle of the target object is determined; the threshold of the target classifier is determined by a receiver operating characteristic curve; the receiver operating characteristic curve is drawn based on multiple eye image samples.
[0124] As an implementation manner, the eye chamber angle state determination module specifically includes:
[0125] a first area ratio determination unit, configured to determine a ratio between an area of a first region in the eye image and an area of a second region in the eye image to obtain an area ratio;
[0126] The first eye chamber angle state acquisition unit is used to input the area ratio into the target classifier to obtain the eye chamber angle state of the target object.
[0127] As an implementation manner, the eye image mentioned above includes a first eye image, a second eye image, a third eye image and a fourth eye image corresponding to the same side eye of the target object;
[0128] Wherein, the first light in the first eye image is above the pupil in the first eye image;
[0129] The first ray in the second ocular image is below the pupil in the second ocular image;
[0130] The first ray in the third ocular image is to the left of the pupil in the third ocular image;
[0131] The first ray in the fourth ocular image is to the right of the pupil in the fourth ocular image;
[0132] Accordingly, the eye chamber angle state determination module 730 specifically includes:
[0133] a second area ratio determination unit, for determining a ratio of an area of a first region to an area of a second region in the first eye image, the second eye image, the third eye image, and the fourth eye image, respectively, to obtain a first area ratio corresponding to the first eye image, a second area ratio corresponding to the second eye image, a third area ratio corresponding to the third eye image, and a fourth area ratio corresponding to the fourth eye image;
[0134] an averaging processing unit, used for averaging the sum of the first area ratio, the second area ratio, the third area ratio and the fourth area ratio to obtain a ratio average value;
[0135] The second eye chamber angle state acquisition unit is used to input the ratio average value into the target classifier to obtain the eye chamber angle state of the target object.
[0136] As an implementation method, the receiver operating characteristic curve is drawn in the following manner, specifically including:
[0137] An image sample acquisition module is used to acquire multiple eye image samples; the multiple eye image samples include positive eye image samples and negative eye image samples; the positive eye image samples are eye images that have been determined to have an open angle; the negative eye image samples are eye images that have been determined to have a closed angle or a suspected closed angle;
[0138] An area ratio determination module, used to determine the ratio of the area of the first area to the area of the second area in each eye image sample, and obtain the area ratio corresponding to each eye image sample;
[0139] A prediction probability acquisition unit, used to input the area ratio corresponding to each eye image sample into the classifier to obtain the prediction probability corresponding to each eye image;
[0140] The curve drawing module is used to draw the receiver operating characteristic curve according to the prediction probability corresponding to each eye image and different preset thresholds.
[0141] As an implementation manner, the area determination module may specifically include:
[0142] An image conversion unit, used for converting an eye image into an eye grayscale image;
[0143] A pixel point acquisition unit is used to perform binarization processing on the eye grayscale image to obtain a plurality of edge pixel points in the eye grayscale image;
[0144] A region determination unit, configured to connect edge pixel points on the first light and edge pixel points on the second light to obtain a first region in the eye image; and connect edge pixel points on the edge of the cornea to obtain a second region;
[0145] The area determination unit is used to determine the number of pixels in the first area of the eye image as the area of the first area of the eye image; and determine the number of pixels in the second area of the eye image as the area of the second area of the eye image.
[0146] The embodiments of the present application also provide corresponding devices and computer-readable storage media for implementing the solutions provided by the embodiments of the present application.
[0147] The device includes a memory and a processor, the memory is used to store a computer program, and the processor is used to execute the computer program so that the device executes the method for determining the state of the eye chamber angle described in any embodiment of the present application.
[0148] The computer-readable storage medium stores a computer program. When the computer program is executed, the device executing the computer program implements the method for determining the state of the eye chamber angle described in any embodiment of the present application.
[0149] The "first" and "second" in the names such as "first" and "second" (if any) mentioned in the embodiments of the present application are only used as name identifiers and do not represent the first or second in order.
[0150] Through the description of the above implementation methods, it can be known that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a readable storage medium, such as a read-only memory (ROM) / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network communication device such as a router) to execute the methods described in each embodiment of the present application or some parts of the embodiments.
[0151] It should be noted that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device embodiment described above is merely schematic, in which the unit described as a separate component may or may not be physically separated, and the component prompted as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative work.
[0152] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A method for determining an eye chamber angle state, characterized in that: The method comprises: Acquire an eye image of a target object; the eye image includes a first light ray and a second light ray; the first light ray is a light ray formed by irradiating the cornea of the target object; the second light ray is a light ray formed by projecting the first light ray through the cornea of the target object onto the iris of the target object and passing through the pupil of the target object; Determine the area of a first region in the eye image, and determine the area of a second region in the eye image; the first region in the eye image is a closed region formed by the first light and the second light on the eye image; the second region in the eye image is a cornea region in the eye image; The state of the eye angle of the target object is determined based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image.
2. The method according to claim 1, characterized in that The determining the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image specifically includes: Based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image and a target classifier, the eye angle state of the target object is determined; the threshold of the target classifier is determined by a receiver operating characteristic curve; the receiver operating characteristic curve is drawn based on multiple eye image samples.
3. The method according to claim 2, characterized in that The determining the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image and a target classifier specifically includes: Determine a ratio of an area of a first region in the eye image to an area of a second region in the eye image to obtain an area ratio; The area ratio is input into the target classifier to obtain the eye chamber angle state of the target object.
4. The method according to claim 2, characterized in that: The eye images include a first eye image, a second eye image, a third eye image and a fourth eye image corresponding to the same side eye of the target object; wherein the first light in the first eye image is above the pupil in the first eye image; The first light in the second eye image is below the pupil in the second eye image; The first light in the third ocular image is to the left of the pupil in the third ocular image; The first light in the fourth ocular image is to the right of the pupil in the fourth ocular image; The determining the eye angle state of the target object based on the ratio of the area of the first region in the eye image to the area of the second region in the eye image and a target classifier specifically includes: Determine a ratio of an area of a first region to an area of a second region in the first eye image, the second eye image, the third eye image, and the fourth eye image, respectively, to obtain a first area ratio corresponding to the first eye image, a second area ratio corresponding to the second eye image, a third area ratio corresponding to the third eye image, and a fourth area ratio corresponding to the fourth eye image; Averaging the sum of the first area ratio, the second area ratio, the third area ratio and the fourth area ratio to obtain a ratio average; The ratio average is input into the target classifier to obtain the eye chamber angle state of the target object.
5. The method according to any one of claims 2 to 4, characterized in that: The receiver operating characteristic curve is drawn in the following manner, specifically including: Acquire the plurality of eye image samples; the plurality of eye image samples include positive eye image samples and negative eye image samples; the positive eye image samples are eye images that have been determined to have an open angle; the negative eye image samples are eye images that have been determined to have a closed angle or a suspected closed angle; In each eye image sample, determining a ratio of an area of the first region to an area of the second region, and obtaining an area ratio corresponding to each eye image sample; Inputting the area ratio corresponding to each eye image sample into a classifier to obtain a prediction probability corresponding to each eye image; The receiver operating characteristic curve is drawn according to the prediction probability corresponding to each eye image and different preset thresholds.
6. The method according to any one of claims 1 to 4, characterized in that: The determining the area of the first region in the eye image and determining the area of the second region in the eye image specifically includes: Converting the eye image into an eye grayscale image; Binarizing the eye grayscale image to obtain a plurality of edge pixel points in the eye grayscale image; Connecting edge pixel points on the first light and edge pixel points on the second light to obtain a first region in the eye image; and connecting edge pixel points on the edge of the cornea to obtain the second region; The number of pixels in the first region of the eye image is determined as the area of the first region of the eye image; and the number of pixels in the second region of the eye image is determined as the area of the second region of the eye image.
7. A device for determining an eye chamber angle state, characterized in that: The device comprises: An image acquisition module, used for acquiring an eye image of a target object; the eye image includes a first light and a second light; the first light is a light formed by irradiating the cornea of the target object; the second light is a light formed by the first light passing through the cornea of the target object, projecting onto the iris of the target object, and passing through the pupil of the target object; an area determination module, used to determine the area of a first area in the eye image and to determine the area of a second area in the eye image; the first area in the eye image is a closed area formed by the first light and the second light on the eye image; the second area in the eye image is a cornea area in the eye image; The eye angle state determination module is used to determine the eye angle state of the target object based on the ratio of the area of the first area in the eye image to the area of the second area in the eye image.
8. The device according to claim 7, characterized in that The eye chamber angle state determination module is specifically used for: determining an eye angle state of the target object based on a ratio of an area of a first region in the eye image to an area of a second region in the eye image and a target classifier; The threshold of the target classifier is determined by a receiver operating characteristic curve; the receiver operating characteristic curve is drawn based on multiple eye image samples.
9. A device for determining an eye chamber angle state, characterized in that: The device comprises a memory and a processor: The memory is used to store a computer program and transmit the computer program to the processor; The processor is used to execute the computer program so that the device performs the steps of the method for determining the eye chamber angle state as described in any one of claims 1-6.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program. When the computer program is executed, a device executing the computer program implements the steps of the method for determining the eye chamber angle state according to any one of claims 1 to 6.