Left and right eye recognition method based on eye ground image features

Abstract

The invention provides a left and right eye recognition method based on eye ground image features, and belongs to the field of medical digital image processing. The method comprises the steps of firstly, processing the acquired eye ground image, then acquiring corresponding information from the processed image via three different methods according to different features of optic discs and blood vessels respectively, finally setting corresponding variables for the three methods and judging whether the methods can be used for judging left and right eyes, performing left and right eye recognition on the combination of the eye ground image according to the recognition result obtained by each method and outputting the recognition result.. The method can be used for quickly and accurately recognizing that the given eye ground image belongs to the left eye or the right eye, is high in recognition accuracy, simple and easy to operate, and has high application value.

Description

technical field [0001] The invention belongs to the field of medical digital image processing, in particular to a method for identifying left and right eyes based on fundus image features. Background technique [0002] Today, with the continuous improvement of people's living standards, the increasing aging of the population and the rapid popularization of smart hardware devices, the prevalence of diabetic retinopathy, glaucoma, high myopia and other diseases has increased significantly, which not only affects people's normal life , but also brought a heavy burden to the family and society. If early detection and early treatment can be achieved, it will undoubtedly be of vital significance to delay the development of the disease, and even allow the vast majority of patients to retain their existing vision, thereby reducing the burden on patients and their families. [0003] Fundus images contain a wealth of unstructured parameters, which is an important diagnostic basis for...

AI Technical Summary

Problems solved by technology

[0008] (1) When a doctor is diagnosing a patient, it is often necessary to use the current fundus image to compare and analyze the fundus image taken by the patient last time or several times before, but the patient will produce multiple fundus images after multiple diagnoses. When observing the left eye (right eye) fundus image, it is necessary to manually search for the left eye (right eye) fundus image from multiple fundus images, which has the disadvantages of heavy workload, low work efficiency, easy fatigue, time-consuming and laborious, etc.;

[0009] (2) Doctors often use image registration technology to observe small changes in blood vessels, optic discs, myopic arcs, and lesions when diagnosing a disease. Image registration also requires two fundus images of the patient in different periods, which still requires manual Disadvantages of manually selecting left-eye (right-eye) fundus images for registration from multiple fundus images

Moreover, when there is a lesion in the patient's eye, due to pathological changes in the affected area, it is likely to affect the doctor's artificial left and right eye recognition and judgment, and delay diagnosis and treatment.

[0010] At present, there is no effective method to identify left and right eyes through fundus images.

Images