Research On Application Of Wavelet Transform And EZW Coding In Medical Image Compression

Wavelet theory has been a topic of research in application math and engineering sciellce.It is an important breakthrouth of mathematics after Fourier transform. JPEG2000 based on wavelet theory are widely applied to many areas, we trys that whether we can applied wavelet theory and compression coding to the field of medical image compression .It goes without saying that medical image information as an important basis for the diagnosis of doctors is very important, our goal is to reserve diagnostic information and compress largely image redundanciesThe paper has deeply studied wavelet theory, innvolving in wavelet selecting, deciding decomposed level in transformed domain and dealling with image of the border, then makes a analysis and comparison for the traditional compression methods based on wavelet theory, such as EZW, SPIHT, SPECK. We makes a detailed analysis of SPIHT algorithm, and points out the shortcomings of the algorithm, such as initialization, determining the importance of subtree coefficient, wavelet coefficients symbols in low-frequency region,then expatiates improved algorithm,carry out compression test. Experiments on standard images and different medical images show that the PSNR of improved algorithm is better than that of original SPIHT at the same ratio. On this basis, we design and develop a system of automatic selection compression algorithm. In the process of the realization of the system,we uses Visual C++.NET as a development tool,Sqlserver 2000 as database,uses object-oriented programming methods,designs medical image compression class,successfully achieve the goal of the gray-scale medical image compression automatically stored.