The Two-Dimensional Genetic Algorithms And Interpolation Algorithms Based On Eno-Haar Wavelet Transform

The thesis is devoted to the application of the ENO(Essentially Non-Oscillatory)wavelet transform to image compression. The algorithms of ENO wavelet is mainly studied, and effectivity and superiority of the algorithms are demonstrated by the simulation experiments. The works are as follows:1 We propose the two-dimensional genetic algorithms of the wavelet coefficients based on the ENO-techniques which have been presented in [1]. Accurately, the multi-orientation ENO-interpolation connection with the characterization of two-dimensional wavelet transform is studied, the weighted mean ENO-algorithm is given and the algorithms are shown by the simulation.2 We present a new method called the Least-Deviation-Smoothness algorithm to eliminate the Gibb's phenomenon based on the related intensity of the wavelet coefficients, and demonstrate the algorithm by the related coefficient analysis as well as the simulation experiments.