Extraction of characteristic points and comparison between the fractal and linear interpolation for the reconstruction of retinal vasculature

Abstract

In this paper we study the compression of the human retina image by the method of Douglas Peucker (DP), and we used fractal interpolation (IFS) and linear interpolation to decompress the coded image. The results show that the method of Douglas-Peucker reduces the image size up to about 92 to 96 %, a rate that is important. The PSNR value obtained by the fractal interpolation is between 27.171 dB and 36.9028 dB, cons, the PSNR value is obtained from linear interpolation 27.659 dB. We see that the PSNR value of fractal interpolation is better than the PSNR value of linear interpolation. We conclude with a fractal interpolation can have a better quality picture, by cons, which is obtained with the interpolation image of lower quality.