Low-illumination color eye fundus image judgment method and enhancement method

Abstract

The invention discloses a low-illumination color eye fundus image judgment and enhancement method. The judgment method comprises the steps that an original color eye fundus image is inputted and converted to an LAB color space; the mean gray value muL and the gray mean square error sigmaL2 of the L channel component of the LAB color space are calculated; and the mean gray threshold mu and the gray mean square error threshold sigma2 of the L channel component are preset, and the image is judged to be a low-illumination eye fundus image if muL<mu and sigmaL2<sigma2. The enhancement method comprises the steps that normalization and gamma correction are performed on the three components of L, A and B of the LAB color space so that three components of Lgamma', Agamma' and Bgamma' are obtained; the Lgamma' component is processed according to a limit contrast self-adaptive histogram equalization method so that an LP' component is obtained; the Agamma' and Bgamma' components are processed so that AP' and BP' are obtained; reverse normalization is performed on LP', AP' and BP'; and the image is converted to an RGB color space from the LAB color space and the image is outputted. The overall brightness of the low-illumination color eye fundus image can be simply and accurately judged and effectively improved.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to a method for judging and enhancing a low-illuminance color fundus image. Background technique [0002] The fundus is the tissue at the back of the eyeball, namely the retina, optic disc, macula, and central retinal artery and vein. Therefore, fundus images can not only directly reflect fundus lesions, but also provide an important basis for the diagnosis of systemic cardiovascular and cerebrovascular diseases. Insufficient illumination and low contrast often occur during the fundus image acquisition process. In addition, the turbidity of the refractive medium of some patients further affects the quality of fundus images, resulting in the appearance of some low-illuminance fundus images, which reduces the efficiency of clinical screening and the accuracy of diagnosis and treatment. Therefore, proposing a method to effectively enhance the brightness and contrast of...

AI Technical Summary

Problems solved by technology

[0003] Existing color fundus image enhancement techniques, such as Contrast Constrained Adaptive Histogram Equalization (CLAHE) method, can adjust image contrast, equalize image grayscale information, enhance details, and provide better visual effects, but for color fundus images and In other words, direct synthesis based on the three channels of red (R), green (G), and blue (B) will produce obvious color cast effect

The guided filtering method can effectively enhance the details of the fundus image without causing color cast, but the image is distorted to a certain extent due to excessive detail enhancement

[0004] The above two methods and other current color fundus image enhancement technologies are mainly aimed at fundus images with general illumination. For the problem of low illumination of fundus images caused by insufficient illumination during the acquisition process and the influence of the patient's own refractive medium, the existing technology does not affect the overall brightness of the image. Adjustments and contrast boosts are poor

Images