A New Method of Fractal IFS Image Database Retrieval Combining Discriminant Function
结合判别函数的分形IFS图像检索

In this paper, a new method of image compare search engine for fractal coding is proposed, namely using image topology characteristic of Iterative Function System (IFS) and applying to image deposit and retrieval. However, the image produces a lot of data by fractal coding and needs an effective method to retrieval, so this paper is based on Fisher discriminant function to estimate images similarity. Experiment shows image compare search engine can get exactly similar image of query image, it also has been testified better for tolerate-error and can effectively improve adaptivity in practice of image compare search engine.