The Investigations for Developing Applied Programs Using Wavelet Software WAVELAB

Wavelets provide very powerful problem-solving tools for analyzing,encoding,compressing,reconstructing,and modeling signals and images.They are useful in capturing,identifying,and analyzing local,multiscale,and nonstationary processes.Wavelet transforms can offer additional insight and performance in data analysis situations where Fourier techniques have previously been used.For example,signals and images with sharp changes and small or irregular detail are often better analyzed with wavelets than with traditional Fourier techniques.With wavelet analysis,engineers and scientists can see and explore aspects of data that other analysis techniques miss,such as trends,breakdown points,discontinuities in higher derivatives,and self-similarity. By now,the amount of wavelets-related software for workstations and PCs,has been constantly multiplying.The scientific computing software such as MATLAB,Mathematica,and Mathcad integrates the wavelet application functions.Also ER Mapper,the most powerful and popular desktop integrated image processing software,provides a component for wavelet compression.Many wavelet analysis toolboxes are widely available in commercial or free of charge.WAVELAB is one of the toolkits,which is developed with MATLAB,an integrated scientific computing environment that combines numeric computation,advanced graphics and visualization,and a high-level programming language.WAVELAB is a library for wavelet analysis,wavelet-packet analysis,cosine-packet analysis and matching pursuit.The library,provided for use in MATLAB 5.X on three platforms:Macintosh,UNIX and Windows NT/98/2000,is available free of charge over the Internet. This paper analyzes the system structure and developing environment on wavelet analysis software WAVELAB,and investigates the problem of developing applied wavelet software based on WAVELAB.An example compressing Lena image by wavelet methods is given,which includes a simple MATLAB program with several results from running the program.The experiment shows that developing applied programs with wavelet software WAVELAB is feasible,getting twice the result with half the effort,and of important significance.