Transforming Retinal Image Analysis: bilinear filter for blood vessel extraction

Abstract:- Accurate segmentation of retinal vasculatures is a vital job required for the identification of a range of clinical disorders. This paper introduces an efficient system for retinal vasculature segmentation. The retinal image undergoes pre-processing through a unique fusion of CLAHE and a homomorphic filter. Bilateral filtering is used in blood vessel segmentation to improve contrast and minimize noise in retinal pictures. Furthermore, a unique thresholding approach called Otsu with entropy thresholding is used for segmentation. At the post-processing phase, the resultant image is generated through the utilization of morphological cleaning and an area-based thresholding approach. The suggested approach’s efficacy is evaluated using the DRIVE and STARE datasets, and its outcomes are benchmarked against current methodologies.