To solve the problem that the image compression time is too long,this paper presented a fractal image encoding method based on convolutional neural network.In this method,we classified the similarities of Range blocks and Domain blocks in fractal image encoding at first,and then used the convolutional neural network to design a new image block classification method,thus achieving the parallel image block classification.Each category searched within its own category during the encoding process,thus reducing some unnecessary matching processes.Finally,the computational complexity of image block matching and the time of image compression were reduced.The experimental results show that the method has higher image restoration quality and shorter compression encoding time,so its appli-cation value is high.
维普
万方数据