Weibull Probability Distribution Function-Based Matched Filter Approach for Retinal Blood Vessels Segmentation

Retinal blood vessels contain an important information that is useful for computer-aided diagnosis of various retinal pathologies such as hypertension, diabetes, glaucoma, etc. Therefore, a retinal blood vessel segmentation is a prominent task. In this paper, a novel Weibull probability distribution function-based matched filter approach is introduced to improve the performance of retinal blood vessel segmentation with respect to prominent matched filter approaches and other matched filter-based approaches existing in literature. Moreover, to enhance the quality of input retinal images in pre-processing step, the concept of principal component analysis (PCA)-based gray scale conversion and contrast-limited adaptive histogram equalization (CLAHE) are used. To design a proposed matched filter, the appropriate value of parameters are selected on the basis of an exhaustive experimental analysis. The proposed approach has been tested on 20 retinal images of test set taken from the DRIVE database and confirms that the proposed approach achieved better performance with respect to other prominent matched filter-based approaches.