Image Enhacement Arithmetic Research Based On Fuzzy Logic And Wavelet Theory

Image enhancement based on wavelet transform and fuzzy theory has been mainly researched in this paper. The algorithm of multi-orientation enhancement based on fuzzy is proposed which confirms the pixel orientation according to rule of the max membership, and then process it. It is fit for low noise image. Then we propose an improved adaptive gaussian filter algorithm carrying convolution between local weighed template and pixels. The weighed coefficients of template are adjusted by changing scale parameter and orientation information given above. This algorithm overcomes the contradiction between de-noising and edge-preserving of adaptive gaussian filter algorithm, but its SNR is less than. the former. The combination algorithm of fuzzy and threshold de-noising is proposed to processes low frequency section by fuzzy filters and high frequency coefficients by soft-threshold and min-max rule. It has a higher SNR and blurry edge. Experiments results indicate that these algorithms have some advantages in de-noising and edge-preserving.