A Novel Retinal Vessel Segmentation Based On Local Adaptive Histogram Equalization

The appearance and structure of blood vessels inretinal images has an important role in diagnosis of diseases. Thispaper proposes a method for segmentation of blood vessels in colorretinal images. We present a method based on preprocessing ofretinal images by enhancing the histogram of grayscale images. Weresearch the effect of linear and non-linear intensity transformationin retinal vessel segmentation. In this paper we focus on twomethods of retinal vessel segmentation including first derivative ofGaussian matched filter and Gaussian matched filter. Thesemethods are used by many researchers [1],[6],[7],[8],[24] but stillthere are some problems in vessel detection. We make use ofadaptive histogram equalization to overcome this drawback. Theproposed technique is tested on publicly available DRIVE database.The effect of improved matched filter can be seen based on accuracycalculation. The accuracy of 0.9353 shows a raise of about 2 percentcompared to existing methods.

کلیدواژه ها:

retinal vessel segmentation ، histogram equalization local adaptive ، matched filter ، first derivative of gaussian