Selective Crypting with Haar-Wavelets.
Alexa Nawotki · 1998
Introduction Cryptographical algorithms usually change all the information included in a data set. Quite often a small part of the information is needed and it would be favourable to transmit that data unchanged. I show a way to construct intermediate versions of the data with gradually reduced information. The critical part of the data, which must not be seen by any other person, can be selected and only this part is changed while the rest stays unchanged. This modified data set can still be used for many applications, but it does not include the real secrets. The central idea for solving this problem is to utilize wavelets: A wavelet-- decomposition or multi-resolution analysis consists of a hierarchy of details, i.e. the original function is split into its smooth and rougher parts with amplitudes of different sizes. Intuitively, little modifications of the original function results from a change of the coefficients belonging to a base with small elongation, and vice versa a