Composition of DCT and Wavelet Transform for Image Compression

The paper reports an advanced image compression algorithm based on a newly proposed standpoint of time-frequency analysis. The advanced algorithm yields 40% improvement against JPEG2000 in compression performance, which means that it attains the same image quality with 40% less data. Compared with JPEG, the improvement may reach 50%. Especially, it makes more improvement at high quality than low quality to the same image. It doubles the coding speed of JPEG2000. But the implementation complexity is only at the level of baseline JPEG. The revolutionary enhancement is critical to high definition (HD) image and video coding technology because of the sea amount of data there. In the paper, the new standpoint of time-frequency analysis is elaborated. Based on the new standpoint, the concept of transform composition is introduced. Furthermore, the composition of DCT and wavelet transform is applied to image compression.