A method and system for biometric measurement of corneal diameter

Abstract

The present invention provides a corneal diameter biometric measurement method and system, which relates to the field of corneal diameter measurement. The measurement method includes obtaining first image information, where the first image information includes an image of the eyeball to be measured of an organism; preprocessing the first image information to obtain the preprocessed first image information, and the first image information is an image of the eyeball to be measured of the organism after denoising; extracting features from the preprocessed first image information to obtain second image information, where the second image information includes the contour features of the cornea to be measured of the organism; sending the second image information to a trained target detection model to obtain a measurement result, and the measurement result includes the diameter information of the cornea to be measured of the organism. By extracting the contour features of the cornea and then detecting the edge points in the corneal contour through the trained target detection model, the present invention effectively realizes non-contact corneal diameter measurement, protects the corneal structure of the detected person, and saves manpower and material resources at the same time.

Description

Technical Field

[0001] The present invention relates to the field of corneal diameter measurement, and more particularly, to a corneal diameter biometric measurement method and system. Background Art

[0002] The accurate measurement of corneal diameter is closely related to the diagnosis and treatment of various diseases, including cataract, corneal refractive surgery, etc. Currently, the measurement of corneal diameter mainly includes manual measurement. Manual measurement requires anesthesia of the eye, and excessive operation time may cause damage to the human cornea. Therefore, there is an urgent need for a non-contact corneal diameter measurement method to achieve non-contact and rapid measurement of corneal diameter. Summary of the Invention

[0003] The purpose of the present invention is to provide a corneal diameter biometric measurement method and system to improve the above problems.

[0004] To achieve the above purpose, the embodiments of the present application provide the following technical solutions:

[0005] On the one hand, the embodiments of the present application provide a corneal diameter biometric measurement method, the method comprising:

[0006] Obtaining first image information, the first image information including an image of the eye to be measured of an organism;

[0007] Preprocessing the first image information to obtain preprocessed first image information, the first image information being an image of the eye to be measured of an organism after denoising;

[0008] Extracting features from the preprocessed first image information to obtain second image information, the second image information including contour features of the cornea to be measured of an organism;

[0009] Sending the second image information to a trained target detection model to obtain a measurement result, the measurement result including diameter information of the cornea to be measured of an organism.

[0010] On the other hand, the embodiments of the present application provide a corneal diameter biometric measurement system, the system comprising:

[0011] An acquisition module, configured to acquire first image information, the first image information including an image of the eye to be measured of an organism;

[0012] A processing module, configured to preprocess the first image information to obtain preprocessed first image information, the first image information being an image of the eye to be measured of an organism after denoising;

[0013] An extraction module for extracting features from the preprocessed first image information to obtain second image information, where the second image information includes the contour features of the cornea to be measured of the organism.

[0014] A sending module for sending the second image information to a trained target detection model to obtain a measurement result, where the measurement result includes the diameter information of the cornea to be measured of the organism.

[0015] In a third aspect, an embodiment of the present application provides a corneal diameter biometric measurement device, which includes a memory and a processor. The memory is used to store a computer program; the processor is used to implement the steps of the above corneal diameter biometric measurement method when executing the computer program.

[0016] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a computer program is stored, and the computer program implements the steps of the above corneal diameter biometric measurement method when executed by a processor.

[0017] The beneficial effects of the present invention are as follows:

[0018] 1. In the present invention, the first image is preprocessed to filter out noise, and then the filtered image is used to extract contour features, effectively extracting the corneal contour of the eye to be measured of the organism from the image. The edge points in the corneal contour are detected by a trained target detection model, and the diameter of the cornea to be measured of the organism is measured by calculating the distance between the edge points. Compared with the manual measurement of the prior art, this method does not require anesthetic treatment for the eye, realizes non-contact corneal diameter measurement, protects the corneal structure of the detected person, and saves manpower and material resources at the same time, providing a fast and intelligent method for measuring the corneal diameter of the organism.

[0019] 2. In the present invention, by improving the rectangular anchor box of the target detection model to a circular anchor box, the detection accuracy of the edge points of the corneal contour is further improved, thereby effectively improving the measurement accuracy of the corneal diameter.

[0020] Other features and advantages of the present invention will be described in the subsequent specification, and part of them will become obvious from the specification, or can be understood by implementing the embodiments of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures specifically pointed out in the written specification, claims, and drawings. Description of the Drawings

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0022] Figure 1 Schematic flow diagram of the corneal diameter biometric method described in the embodiments of the present invention.

[0023] Figure 2 Schematic structural diagram of the corneal diameter biometric system described in the embodiments of the present invention.

[0024] Figure 3 Schematic structural diagram of the corneal diameter biometric device described in the embodiments of the present invention.

[0025] Reference numerals in the figure: 901, acquisition module; 902, processing module; 903, extraction module; 904, sending module; 9021, preset unit; 9022, clustering unit; 9023, first processing unit; 9024, second processing unit; 9031, third processing unit; 9032, fourth processing unit; 9033, fifth processing unit; 90311, first acquisition unit; 90312, sixth processing unit; 90313, seventh processing unit; 9041, second acquisition unit; 9042, eighth processing unit; 9043, training unit; 9044, ninth processing unit; 9045, tenth processing unit; 9046, eleventh processing unit; 800, corneal diameter biometric device; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. Detailed implementation manners

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed present invention, but merely represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0027] It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present invention, terms such as "first", "second", etc. are only used for distinguishing descriptions and cannot be construed as indicating or implying relative importance.

[0028] Embodiment 1:

[0029] This embodiment provides a method for biometric measurement of corneal diameter. It can be understood that in this embodiment, a scenario can be set up. For example, an ultrasonic biomicroscope is used to observe the eyeball of the subject to be measured, and a screenshot is taken to obtain an ultrasonic biomicroscopic image of the eyeball of the subject to be measured, and the diameter measurement of the cornea of the eyeball to be measured is realized according to the screenshot.

[0030] See Figure 1 , the figure shows that this method includes step S1, step S2, step S3 and step S4, where:

[0031] Step S1: Obtain first image information, where the first image information includes an image of the eyeball of the subject to be measured;

[0032] It can be understood that the first image information in this step is an ultrasonic biomicroscopic image obtained by observing the eyeball of the subject to be measured with an ultrasonic biomicroscope and taking a screenshot, or an OCT image corresponding to the eyeball is obtained by observing the eyeball with an optical coherence tomography scanner (OCT), so as to realize the diameter measurement of the cornea of the eyeball to be measured.

[0033] Step S2: Preprocess the first image information to obtain the preprocessed first image information, where the first image information is an image of the eyeball of the subject to be measured after denoising;

[0034] In this step, denoising the ultrasonic biomicroscopic image of the eyeball of the subject to be measured obtained by taking a screenshot is beneficial to improving the subsequent extraction of corneal contour features and ensuring the accuracy of the corneal contour.

[0035] It can be understood that step S2 further includes step S21, step S22, step S23 and step S24, where:

[0036] Step S21: Preset a first clustering center and a second clustering center, where the first clustering center is the pixel point corresponding to the eyeball image, and the second clustering center is the pixel point corresponding to the noise point;

[0037] It can be understood that the first clustering center of the pixel points corresponding to the eye image is represented by the first feature vector, and the second clustering center of the pixel points corresponding to the noise points is represented by the second feature vector, where the noise points are the points discrete from the eye image in the first image information.

[0038] Step S22: Perform clustering analysis on the first image information by using the first clustering center and the second clustering center to obtain first information, where the first information includes the clustering result of the first clustering center and the clustering result of the second clustering center;

[0039] It can be understood that by sending both the first feature vector and the second feature vector to the K-means clustering model for clustering analysis, the clustering result of the first clustering center and the clustering result of the second clustering center can be obtained.

[0040] It should be noted that when both the first feature vector and the second feature vector are sent to the K-means clustering model for clustering analysis, the effectiveness of the clustering result can be judged by a criterion function, where the criterion function is specifically:

[0041]

[0042] In the above formula, C i is the center point corresponding to each cluster, μ i is the mean value corresponding to each cluster, that is, by calculating the squared error between the data in the feature space of the center points of each cluster and the mean value corresponding to each cluster, the criterion function can be obtained, and when the Z value is the smallest, a clustering result is generated. The specific calculation formula of μ i is:

[0043]

[0044] In the above formula, N i is the number of data in the i-th cluster, x is the data in the feature space,

[0045] Step S23: Filter out the clustering result of the second clustering center included in the first information to obtain second information;

[0046] Step S24: Perform dilation processing and erosion processing on the second information by using convolution sum to obtain the preprocessed first image information.

[0047] It can be understood that by performing dilation processing and erosion processing on the second information, a continuous and clear image can be obtained.

[0048] Step S3: Extract features from the preprocessed first image information to obtain second image information, where the second image information includes the contour features of the cornea to be measured of the organism;

[0049] It is understandable that step S3 further includes step S31, step S32, and step S33, where:

[0050] Step S31: Send the preprocessed first image information to a convolutional neural network for feature extraction to obtain a first feature image;

[0051] It is understandable that the convolutional neural network includes 3 3X3 convolutional layers. Feature information is extracted through the 3 3X3 convolutional layers to obtain a first feature image.

[0052] It is understandable that before step S31, there are also step S311, step S312, and step S313, where:

[0053] Step S311: Obtain a reference image, where the reference image is the corneal contour after manual segmentation;

[0054] Step S312: Send the reference image and the processed first image information to an attention mechanism to obtain the weights of the features in the preprocessed first image information;

[0055] Step S313: Weight the corresponding features in the preprocessed first image information according to the weights of the features in the preprocessed first image information.

[0056] In this embodiment, when extracting and segmenting corneal contour features, due to unclear extraction of important feature information of the image, pseudo-contours are easily segmented. Therefore, an attention mechanism is added to this neural network model, which can extract specified features. By modeling the features of the manually labeled segmented corneal contour image and performing weighted summation of the trained weights and the corresponding features, efficient extraction of important contour features can be achieved. In addition, since the attention mechanism is introduced into the neural network model, the loss function of the neural network model needs to be improved. Specifically:

[0057] L = μ1L1 + μ2L2 + μ3L3 + μ4L4

[0058] In the above formula, μ1, μ2, μ3, and μ4 are the weight coefficients of each cross-entropy loss respectively, and L1, L2, L3, and L4 are the losses between the feature maps after the first, second, third, and fourth upsamplings, that is, the second image information, and the manually labeled segmented corneal contour image.

[0059] Step S32: Send the first feature image to a pooling layer for downsampling to obtain a second feature image, where the second feature map is a feature image with reduced dimensions;

[0060] It can be understood that by performing downsampling layer by layer through a 2X2 pooling layer, the dimension of the feature map is reduced, and at the same time, some other unimportant feature information is filtered out. After multiple convolutions and poolings, the high-level features of the image can be effectively extracted.

[0061] Step S33: Upsample the second feature image using deconvolution to obtain second image information.

[0062] It can be understood that by performing upsampling step by step through 2X2 deconvolution, the dimension of the feature image is restored, and the feature information in the second feature image is gradually restored, realizing the mapping of low-resolution features to the segmentation map.

[0063] Step S4: Send the second image information to the trained object detection model to obtain a measurement result, where the measurement result includes the diameter information of the cornea to be measured of the organism.

[0064] It can be understood that before step S4, there are steps S41, S42, and S43, where:

[0065] Step S41: Obtain preset anchor box parameters, and the anchor box parameters are circular anchor boxes;

[0066] It can be understood that in this embodiment, the selected object detection model is YOLOV5. By modifying the anchor box parameters in the original algorithm to obtain the parameters of circular anchor boxes, the accuracy of object detection can be effectively improved.

[0067] Step S42: Establish a loss function according to the feature information of the circular anchor box, where the feature information includes the shape feature of the circular anchor box;

[0068] It can be understood that a loss function is established according to the feature information of the circular anchor box, specifically:

[0069]

[0070] In the above formula, IoU is the intersection over union of the predicted box and the ground truth box, S A is the area of the predicted box, S B is the area of the ground truth box, S C is the area of the smallest box containing box A and box B. Since the original rectangular anchor box is changed to a circular anchor box, the formula for calculating the overlap degree of the predicted box and the ground truth box needs to be improved, specifically:

[0071]

[0072] In the above formula, IoU is the intersection over union of the predicted bounding box and the ground truth bounding box, that is, the overlap degree. It should be noted that when IoU is 0, it means that there is no intersection between bounding box A and bounding box B, that is, the two boxes are very far apart; when IoU is 1, it means that bounding box A and bounding box B completely overlap.

[0073] Step S43: Train the object detection model according to the anchor box parameters and the loss function to obtain a trained object detection model.

[0074] It should be noted that training the YOLOV5 object detection model using anchor box parameters and loss function is a technical solution well-known to those skilled in the art, so it will not be elaborated in this application.

[0075] It can be understood that step S4 further includes steps S44, S45, and S46, where:

[0076] Step S44: Send the second image information to the trained object detection model to obtain third information, where the third information is the position information of the circular anchor boxes detecting two edge points of the cornea to be measured of the organism.

[0077] It can be understood that the trained object detection model can effectively detect two edge points of the cornea to be measured of the organism. Compared with the rectangular anchor boxes in the prior art, the present application effectively improves the accuracy of edge point detection.

[0078] It should be noted that the edge points include at least one upper edge point and at least one lower edge point of the cornea detection area. The diameter of the cornea can be calculated by the upper and lower two edge points corresponding to the cornea detection area obtained by detection.

[0079] Step S45: Obtain the center coordinates corresponding to the circular anchor boxes according to the position information of the circular anchor boxes of the two edge points of the cornea to be measured of the organism.

[0080] Step S46: Calculate the diameter information of the cornea to be measured of the organism according to the circular coordinates corresponding to the two circular anchor boxes.

[0081] It can be understood that the distance between the circular coordinates corresponding to the two circular anchor boxes, that is, the diameter information of the cornea to be measured of the organism, can be calculated through the distance formula, specifically:

[0082]

[0083] In the above formula, d is the diameter of the cornea to be measured of the organism, x1 and x2 are the abscissas corresponding to the center coordinates of two circular anchor boxes respectively, and y1 and y2 are the ordinates corresponding to the center coordinates of two circular anchor boxes respectively. It should be noted that by calculating the distances between the upper edge points and each lower edge point, the maximum value among them is selected as the first diameter information of the cornea to be measured of the organism. Then, by replacing the upper edge point and calculating the distances between it and each lower edge point again, the maximum value among them is selected as the second diameter information of the cornea to be measured of the organism. By repeating the above calculation process, multiple diameter information of the corneas to be measured of the organism can be obtained. By summing up and taking the average of the multiple diameter information of the corneas to be measured of the organism, more accurate diameter information of the corneas to be measured of the organism can be obtained. It can be understood that the more upper edge points are selected, the more accurate the calculated diameter information of the corneas to be measured of the organism will be.

[0084] Embodiment 2:

[0085] As Figure 2 shown, this embodiment provides a corneal diameter biometric measurement system, which includes an acquisition module 901, a processing module 902, an extraction module 903, and a sending module 904, where:

[0086] The acquisition module 901 is used to acquire first image information, and the first image information includes an image of the eyeball to be measured of the organism;

[0087] The processing module 902 is used to preprocess the first image information to obtain preprocessed first image information, and the first image information is an image of the eyeball to be measured of the organism after denoising;

[0088] The extraction module 903 is used to extract features from the preprocessed first image information to obtain second image information, and the second image information includes the contour features of the cornea to be measured of the organism;

[0089] The sending module 904 is used to send the second image information to the trained target detection model to obtain a measurement result, and the measurement result includes the diameter information of the cornea to be measured of the organism.

[0090] In a specific implementation manner of the present disclosure, the processing module 902 includes a preset unit 9021, a clustering unit 9022, a first processing unit 9023, and a second processing unit 9024, where:

[0091] The preset unit 9021 is used to preset a first clustering center and a second clustering center. The first clustering center is the pixel point corresponding to the eyeball image, and the second clustering center is the pixel point corresponding to the noise point;

[0092] A clustering unit 9022, configured to perform clustering analysis on the first image information by using the first clustering center and the second clustering center to obtain first information, where the first information includes the clustering results of the first clustering center and the clustering results of the second clustering center;

[0093] A first processing unit 9023, configured to filter out the clustering results of the second clustering center included in the first information to obtain second information;

[0094] A second processing unit 9024, configured to perform dilation processing and erosion processing on the second information by using convolution sum to obtain preprocessed first image information.

[0095] In a specific embodiment of the present disclosure, the extraction module 903 includes a third processing unit 9031, a fourth processing unit 9032, and a fifth processing unit 9033, where:

[0096] The third processing unit 9031 is configured to send the preprocessed first image information to a convolutional neural network for feature extraction to obtain a first feature image;

[0097] The fourth processing unit 9032 is configured to send the first feature image to a pooling layer for downsampling to obtain a second feature image, and the second feature map is a feature image with reduced dimensions;

[0098] The fifth processing unit 9033 is configured to perform upsampling on the second feature image by using deconvolution to obtain second image information.

[0099] In a specific embodiment of the present disclosure, before the third processing unit 9031, there are a first acquisition unit 90311, a sixth processing unit 90312, and a seventh processing unit 90313, where:

[0100] The first acquisition unit 90311 is configured to acquire a reference image, and the reference image is a cornea contour after manual segmentation;

[0101] The sixth processing unit 90312 is configured to send the reference image and the processed first image information to an attention mechanism to obtain the weights of the features in the preprocessed first image information;

[0102] The seventh processing unit 90313 is configured to weight the corresponding features in the preprocessed first image information according to the weights of the features in the preprocessed first image information.

[0103] In a specific embodiment of the present disclosure, before the sending module 904, there are a second acquisition unit 9041, an eighth processing unit 9042, and a training unit 9043, where:

[0104] A second acquisition unit 9041, configured to acquire preset anchor box parameters, where the anchor box parameters are circular anchor boxes;

[0105] An eighth processing unit 9042, configured to establish a loss function according to the feature information of the circular anchor box, where the feature information includes the shape feature of the circular anchor box;

[0106] A training unit 9043, configured to train the object detection model according to the anchor box parameters and the loss function to obtain a trained object detection model.

[0107] In a specific implementation manner of the present disclosure, the sending module 904 further includes a ninth processing unit 9044, a tenth processing unit 9045, and an eleventh processing unit 9046, where:

[0108] The ninth processing unit 9044 sends the second image information to the trained object detection model to obtain third information, where the third information is the position information of the circular anchor boxes detecting two edge points of the cornea to be measured of the organism;

[0109] The tenth processing unit 9045 obtains the center coordinates corresponding to the circular anchor box according to the position information of the circular anchor boxes of the two edge points of the cornea to be measured of the organism;

[0110] The eleventh processing unit 9046 calculates the diameter information of the cornea to be measured of the organism according to the circular coordinates corresponding to the two circular anchor boxes.

[0111] It should be noted that for the system in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0112] Embodiment 3:

[0113] Corresponding to the above method embodiment, a corneal diameter biometric device is further provided in this embodiment. A corneal diameter biometric device described below can be mutually corresponded and referred to with a corneal diameter biometric method described above.

[0114] Figure 3 It is a block diagram of a corneal diameter biometric device 800 shown according to an exemplary embodiment. As Figure 3 shown, the corneal diameter biometric device 800 may include: a processor 801, a memory 802. The corneal diameter biometric device 800 may further include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0115] Among them, the processor 801 is used to control the overall operation of the corneal diameter biometric device 800 to complete all or part of the steps in the above corneal diameter biometric method. The memory 802 is used to store various types of data to support the operation of the corneal diameter biometric device 800. Such data may include, for example, instructions for any application or method operating on the corneal diameter biometric device 800, as well as application-related data, such as contact data, sent and received messages, pictures, audio, video, and so on. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, a magnetic disk, or an optical disc. The multimedia component 803 may include a screen and an audio component. Among them, the screen may be a touch screen, for example, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone, and the microphone is used to receive external audio signals. The received audio signal may be further stored in the memory 802 or sent through the communication component 805. The audio component further includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above other interface modules may be a keyboard, a mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the corneal diameter biometric device 800 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, near field communication (NFC), 2G, 3G, or 4G, or a combination of one or more of them. Accordingly, the communication component 805 may include: a Wi-Fi module, a Bluetooth module, and an NFC module.

[0116] In an exemplary embodiment, the corneal diameter biometric device 800 can be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components, and is used to execute the above-mentioned corneal diameter biometric method.

[0117] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When the program instructions are executed by a processor, the steps of the above-mentioned corneal diameter biometric method are implemented. For example, the computer-readable storage medium can be the above-mentioned memory 802 including program instructions, and the above program instructions can be executed by the processor 801 of the corneal diameter biometric device 800 to complete the above-mentioned corneal diameter biometric method.

[0118] Embodiment 4:

[0119] Corresponding to the above method embodiment, a readable storage medium is also provided in this embodiment. A readable storage medium described below can be correspondingly referred to with a corneal diameter biometric method described above.

[0120] A readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the corneal diameter biometric method in the above method embodiment are implemented.

[0121] The readable storage medium can specifically be various readable storage media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store program codes.

[0122] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0123] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for biometric measurement of corneal diameter, characterized in that, Comprising: Obtain first image information, where the first image information includes an image of the eyeball to be measured of an organism; Preprocess the first image information to obtain preprocessed first image information, where the first image information is an image of the eyeball to be measured of an organism after denoising; Extract features from the preprocessed first image information to obtain second image information, where the second image information includes the contour features of the cornea to be measured of an organism; Send the second image information to a trained target detection model to obtain a measurement result, where the measurement result includes the diameter information of the cornea to be measured of an organism; Preprocess the first image information to obtain preprocessed first image information, including: Preset a first clustering center and a second clustering center, where the first clustering center is the pixel point corresponding to the eyeball image, and the second clustering center is the pixel point corresponding to the noise point; Perform clustering analysis on the first image information using the first clustering center and the second clustering center to obtain first information, where the first information includes the clustering result of the first clustering center and the clustering result of the second clustering center; Filter out the clustering result of the second clustering center included in the first information to obtain second information; Perform dilation processing and erosion processing on the second information using convolution sum to obtain preprocessed first image information; Extract features from the preprocessed first image information to obtain second image information, including: Send the preprocessed first image information to a convolutional neural network for feature extraction to obtain a first feature image; Send the first feature image to a pooling layer for downsampling to obtain a second feature image, where the second feature image is a feature image with reduced dimensions; Perform upsampling on the second feature image using deconvolution to obtain second image information; Before sending the second image information to a trained target detection model to obtain a measurement result, including: Obtain preset anchor box parameters, where the anchor box parameters are circular anchor boxes; Establish a loss function according to the feature information of the circular anchor box, where the feature information includes the shape features of the circular anchor box; Train the target detection model according to the anchor box parameters and the loss function to obtain a trained target detection model.

2. The corneal diameter biometric method according to claim 1, characterized in that Before sending the preprocessed first image information to a convolutional neural network for feature extraction to obtain a first feature image, including: Obtain a reference image, where the reference image is the cornea contour after manual segmentation; Send the reference image and the preprocessed first image information to an attention mechanism to obtain the weights of the features in the preprocessed first image information; Weight the corresponding features in the preprocessed first image information according to the weights of the features in the preprocessed first image information.

3. A corneal diameter biometric measurement system, characterized in that, For implementing the method as claimed in claim 1, including: An acquisition module, configured to acquire first image information, where the first image information includes an image of the eyeball to be measured of an organism; A processing module, configured to preprocess the first image information to obtain preprocessed first image information, where the first image information is an image of the eyeball to be measured of an organism after denoising; An extraction module, configured to extract features from the preprocessed first image information to obtain second image information, where the second image information includes the contour features of the cornea to be measured of the organism; A sending module, configured to send the second image information to a trained target detection model to obtain a measurement result, where the measurement result includes the diameter information of the cornea to be measured of the organism; The processing module includes: A preset unit, configured to preset a first clustering center and a second clustering center, where the first clustering center is the pixel point corresponding to the eye image, and the second clustering center is the pixel point corresponding to the noise point; A clustering unit, configured to perform clustering analysis on the first image information by using the first clustering center and the second clustering center to obtain first information, where the first information includes the clustering result of the first clustering center and the clustering result of the second clustering center; A first processing unit, configured to filter out the clustering result of the second clustering center included in the first information to obtain second information; A second processing unit, configured to perform dilation processing and erosion processing on the second information by using convolution sum to obtain the preprocessed first image information; The extraction module includes: A third processing unit, configured to send the preprocessed first image information to a convolutional neural network for feature extraction to obtain a first feature image; A fourth processing unit, configured to send the first feature image to a pooling layer for downsampling to obtain a second feature image, where the second feature image is a feature image with reduced dimensions; A fifth processing unit, configured to perform upsampling on the second feature image by using deconvolution to obtain second image information; Before the sending module, it includes: A second obtaining unit, configured to obtain preset anchor box parameters, where the anchor box parameters are circular anchor boxes; An eighth processing unit, configured to establish a loss function according to the feature information of the circular anchor box, where the feature information includes the shape features of the circular anchor box; A training unit, configured to train the target detection model according to the anchor box parameters and the loss function to obtain a trained target detection model.

4. The corneal diameter biometric measurement system according to claim 3, characterized in that, Before the third processing unit, it includes: A first obtaining unit, configured to obtain a reference image, where the reference image is the cornea contour after manual segmentation; A sixth processing unit, configured to send the reference image and the preprocessed first image information to an attention mechanism to obtain the weights of the features in the preprocessed first image information; A seventh processing unit, configured to weight the corresponding features in the preprocessed first image information according to the weights of the features in the preprocessed first image information.

Images