Edge detection algorithm based on AIVHE for retinal images

The goal of segmentation is to represent an image more easily & meaningfully. Edge detection method used is an important property of extracting image characteristics in image segmentation, identification and analysis. In this paper we proposed a new algorithm called SUSAN edge detection algorithm which improves the performance of the traditional test edge detection operators and has good segmentation accuracy. Pre-processing techniques are also applied to obtain better results. The proposed algorithm first filters the noisy images and enhances the contrast of the image by AIVHE method. Here we have studied the effect of the three different histogram equalization methods namely the traditional HE, CLAHE and AIVHE for certain performance parameters on medical retinal images. Susan edge detection gives better performance in presence of noise and provides good edge, connectivity, better localization and no false edges compare to other derivative based operators.