Fractal image compression using neural networks

We propose the application of neural network technologies to fractal image compression and decompression by implementing numerous comparisons and transformations that are required by the traditional methods. In this way, the computation time of encoding can be reduced for image compression, and the compression/decompression process can be executed in parallel. Simulated results show that the neural network approach can obtain a high compression-ratio and a clear decompressed image. Thus, it validates the effectiveness of our neural network approach.