A Proposed Method for Contour Extraction of an Image Based On Self-Affine Mapping System by Fractal Coding

A self-affine mapping system instead of the energy minimization procedure is used to approach and fit the roughly drawn line to the object contour. The self-affine map’s parameters are detected by analyzing the block wise self-similarity of a gray-scale image using an algorithm in fractal encoding techniques. The edges are attracting mapping points in self-affine mapping is utilized in the proposed method. The proposed method accurately produces smooth curves sharp corners and to extract both distinct edges and blurred edges. The large gaps between the hand-draw line and the contour can be fitted well by our proposed method and algorithms. In this method block size is progressively decreased. These procedures reduce the time required for drawing contours by hands. As a result, highly accurate contours were extracted in our proposed method.