Design And Implication Of Image Compression System Based On Wavelet And DM6446

In recent years, image compression technology has a wide range of applications such as digital TV, video telephony, video conferencing, telemedicine and many other areas. With the rapid development of Internet and mobile communication technology. the demands of image information continues to grow and the research on image compression technology becomes one of the hot and widespread concern topics. Nowadays, internationally, there are some standard algorithms (Such as JPEG, JPEG2000, H.263, H.264, MPEG2, MPEG4 and other international standard algorithms), these algorithms make use of block DCT transformation as the core algorithm. With the continuous expansion of the scope of application, original algorithm gradually exposes the shortcomings and is unable to meet the people needs of high-quality images. At the same time, the existed Image compression systems have been mostly implemented in large-scale integrated circuit and have complex implementation process, high cost and limited applications.With the development and maturation of wavelet theory, Wavelet transform combines with image compression technology, the good time and frequency localization,the flexible analysis capabilities of non-stationary image signals and the adaptability of Human visual features overcome the shortcomings of the original compression algorithm effectively.At the same time,the embedded system has lots of advantages such as the hardware and software can be cut, high reliability and small volume, therefore, compression algorithm in my paper was implemented on an embedded platform, the one hand, the development cycle of system can be shortened and the costs can be reduced; On the other hand, it can effectively expand the scope of application of image compression system.In this paper, firstly, we described the principle of wavelet transform and the dual-core platform architecture DM6446 and choose improved wavelet transform—lifted wavelet transform algorithm transformed the image, the results were coded by the embedded wavelet transform zero tree and we achieved image compression. Then we transplanted the compression algorithm to DM6446 and realized image compression system. Finally, we tested the system. The results of a large number of images showed that: this system has highly integration, low cost and good result of image compression and can meet the requirements of practical applications.