Study For SPIHT Compression Arithmetic And Application Based On Lifting Wavelet Transform

With the rapid development of the computer and communication technology, image application is getting widely used. Man's demand for information doesn't barely confine to traditional pronunciation, characters message , but the high-quality multimedia information that incorporates video, image, sound and characters into an organic whole, but it demands highly on band width of signal transmission and data memory capacity, etc .. With the rapid development of network and communication and extensive use of multimedia technology, the copyright protection of digital median work is becoming more and more important. Digital watermarking is viewed as an effective way of intellectual property right protection of multimedia data.So,set about from method of image data compression, the paper introduces wavelet and wavelet transform firstly. Wavelet transform is an analytical method of time-scalable, which has the characteristics of multi-resolution, is partly transformed in space and frequency, and can realize self-adaptation in image analysis, But must carry on the huge convolution operation, it is complex. Mallat algorithm and the theory and methods of embedded image coding are introduced in this paper. The Set Partitioning Hierarchical Trees (SPIHT)algorithm is presented in detail, which is based on the zerotree structure.Then,the paper analyzes the lifting scheme of wavelet transform and how to transform image with it. Compares with the traditional wavelet transformation, the wavelet transformation of lifting scheme have some characteristics: the operating speed more quickly, home position operation, the DSP chip realization. The paper introduces the theory of lifting and its key techniques, especially, analyzing the whole process for factoring the wavelet transform into lifting steps, such as integer wavelet transforms.And then, in view of the problems of complicated convolution process of wavelet transform, repeated calculations and a large number of needed memories of SPIHT algorithm after wavelet transformation, a new image compression method based on improved SPIHT and integer lifting wavelet transform is proposed. the method not only considers human visual specialities to improve quality of restructured photo but also introduces the idea of max value table to save time of index.There is the characteristic of economical memory ,fast computation and simple encode.The peak SNR for reconstruction image also has some improvement. Experiments demonstrate that it is a viable and effective image compression method.Finally, Image compression based on integer lifting wavelet transform and improved SPIHT is applied in watermark, blind digital watermarking algorithm based on integer lifting wavelet transform. Quantity of experiment data shows that algorithms have good robustness, and the imperceptibility of watermark can be ga uaranteed at the same time. To sum up, three characteristics are achieved in this algorithm .First, secret message is directly embedded in a improve-SPIHT, when the appropriate minbit is choosed, the contradiction between the image quality of the stego-image and the hiding capacity of the cover image is solved possibly and the hidden secret message ia extracted progressively during the improve-SPIHT decoding process. Second, the security of the watermarking algorithm is good, when there is extracted secret message successfully ,there are known present minbit , correspondingτP,secret matrix S, and the interger weight matrix W; Third, have very large embedding capacity, it allows the size of secret image to be the same as the cover image.