March 17th, 2002
Part of my project for the Wavelet course involves representing a 2-D image file by wavelet coefficients. The results shown are for the simple Haar wavelets, but the code is easily modified to include other forms of wavelets and scaling function; in fact, I hope to implement B-splines in the near future if time permits. The selling feature of why one would want to use wavelets is evident in the compression routine developed. Although it doesn't really save much space in the "matlab environment" (because matlab stores integers as real numbers), one notices that we have a fairly good reconstruction even when only 10% of the coefficients are used. i.e., instead of representing a 256x256 picture as an array of 65536 real numbers, we only need to represent it as 6554 real numbers and 13108 integers. In any case, please feel free to take a look at the code, workbook and project writeup (when available =) and please reference anything that you found useful.
cheers
ben