Segmenting Retinal Vessels with a Multi-scale Modified Dolph-Chebyshev Type I Function Matched Filter

Abstract

In this paper, a new algorithm for retinal vessels segmentation is proposed. The algorithm is based on a multiscale modified Dolph-Chebyshev type I function matched filter. Fundus images from the DRIVE and STARE databases are utilized to evaluate the performance of the proposed algorithm. Several performance indicators, such as specificity, sensitivity, and accuracy are used for performance evaluation. Experimental results show that for the DRIVE database, the results of the proposed algorithm are superior to those produced by all compared algorithms. When tested on pathological images from the STARE database, the proposed algorithm also performs better than all competing methods. This indicates that the proposed algorithm is suitable to be used in automatic retinal diseases diagnosis tools.