Color Image Compression Coding Based On Multi-band Wavelet Transform

Image compression coding technology is basically a hybrid coding based on transformation,so mathematical transformation is the core and key of image compression coding.The structure of mathematical transformation determines the complexity of image compression algorithm,including its operation efficiency,coding efficiency,algorithm parallelism and so on.The static image compression standard JPEG and the new generation image compression standard JPEG2000 adopt two different mathematical transformation structures.JPEG adopts the DCT as the core transformation algorithm,while JPEG2000 adopts the 2-band wavelet transform as the core transformation algorithm.There is a big difference between the two kinds of mathematical transformation,which leads to the different characteristics of compression coding algorithm.The former algorithm is concise and parallel,but has block effect and low coding efficiency,while the latter has high coding efficiency,but the algorithm is complex and the cost of hardware-based implementation is high.In order to solve these problems,this paper studies the multi-band wavelet transform compression coding algorithm for color image based on the theory of multi-band wavelet transform,and makes a thorough discussion on the implementation details of the color image compression coding algorithm.Multiband wavelet transform not only has better energy concentration than DCT in theory,but also can be designed as overlapping orthogonal transform algorithm when designing software and hardware algorithms in practice,which can greatly reduce hardware implementation complexity and cost.Experiments show that the reconstructed image based on multi-band wavelet transform is better than JPEG compression standard in objective PSNR and subjective perception.