The Research Of Remote Sensing Image Compression Based On Wavelet Zero-tree Coding

Remote sensing image is a dominating form to get and represent remote sensing data. As the desire for speed and amount in data capture is increasing, real-time image compression method and platform becomes an important developing orientation for remote sensing system. Wavelet algorithm is the revolutionary evolution result of Fourier analysis, which has dual-significance for its profound theory and broad application.This thesis summarized and compared traditional image compression methods, and expatiated on still image compression and decompression based on the algorithm of wavelet transform. Images are transformed firstly from spatial domain to frequency domain through wavelet transform. Because of the multi-resolution property of wavelet transform, original images will be transformed into a certain number of sub-images in different resolution. Among them, the ones in lower resolution comprising most energy of the original image are regarded as approximations to it, while the ones in higher resolution embody details of the original. This transform is usually called image wavelet decomposition. Meanwhile, these sub-images together can be recovered into the original image by reverse wavelet transform, which is called wavelet image reconstruction.With analysis on characteristics of wavelet coefficient, this thesis introduced the superiority of wavelet transform in image coding. Beginning with the questions which should be considered in image compression using wavelet transform, the thesis discussed the property, amount of decomposing layers and boundary processing of different wavelet bases, and the conversion between RGB images and YUV images. According to the distribution of transform coefficient, the thesis analyzed the characteristics of embedded Zero-tree coding and provided the concept, theory and procedure of Zero-tree coding. After that, a series of experiments were carried out to compress the same image in different compression ratios, and gave a subjective evaluation on the recovered images. In image compression, it is an important study to evaluate the quality of the reconstructed image, in which the lack of objectivity and accuracy in the evaluation method may mislead the further research of image...