The Research Of Image Information Compression Techniques

Because of the difficulty in transmission and storage of mass data of digital image, people considered how to raise the compression rate of the picture possibly under the certain picture quality. In this paper, The algorithms of image compression based on BP neural network are studied mainly, and the BP neural network based on quadtree fracture coding is proposed later on.In sum, this paper is studied according to the following hierarchy and research mechanism:1. Use of the SNR and compression rate as a standard, this paper firstly discuss BP algorithm and explode the mechanism of image compression based on BP neural network. Then we make a study of the key technology in the application and several kinds of learning rules are used to compress the static image. After a series of experiments are executed, we analyze and summarize the relationship between the compression performance and the parameters of BP neural network. From above analysis,we put forward several method as followed: formulary about study rate η,weight Δω,choice about Sigmoid function and studied on hierarchical BP network. Realized the performance analysis and picture rebuild at last.2. Proposed the neural network based on quadtree image coding through the analysis about quadtree fracture coding, given the algorithms of arithmetic and realized the program with the stack method. Analytic results shows that this algorithm has shorter compressing time and lower arithmetic complexity.