A New Method to Segment Retinal Vessel Based on Gabor Wavelet

Abstract

Retinal vessel segmentation is important for better analysis and diagnosis of vascular disease. This paper proposed a novel automatic retinal segmentation approach based on Gabor wavelet. First, the green channel is extracted from a retinal fundus image, then, the method use Gabor wavelet transformation to enhance the contrast of the vessel, finally, a shelf adaptive threshold segmentation technology is used for vessel extraction. The performance of the approach is estimated by two publicly retinal fundus image databases DRIVE and STARE. The results show that the new approach obtains better segmentation effect.