The Study Of Digital Image Sharpness Based On Wavelet Analysis

Wavelet analysis has a certain effect on digital image denoising as a better method of transform domain. But besides noises, there are some other interference factors which weaken the vision effect of digital image greatly, like image contrast, image brightness and image sharpness, etc. So the vision effect of digital image can't be improved well just by image denoising, and we need an mtegrated method which combines many digital image processing methods.In the thesis, the author firstly clears up some basic theories of wavelet analysis, revises some formulae and proves and deduces some theorems and formulae in detail and gives a counter example and paradox. And then through the study of many digital image processing methods, we select image denoising and enhancement based on wavelet analysis, gray image histogram modification and image smoothness based on median filter. Moreover, we develop an integrated method of digital image sharpness on the basis of the above four methods. The integrated method accords to the following steps: blurred noisy imageâ†'image denoising based on wavelet analysis— image histogram modificationâ†'image enhancement based on wavelet analysisâ†' image smoothness based on median filterâ†'sharpness image. The result of blurred noisy image processing proves that the integrated method has a good image vision effect and sharpness. So the conclusion we can come to from above all is that the integrated method is an effective and feasible method to improve the sharpness of blurred noisy image. In the thesis, the gray histogram of blurred noisy image has characteristics of about normal distribution, while the gray histogram of sharpness image has characteristics of skewed distribution.From the study of above digital image processing methods, we can come to the conclusion that the integrated method of digital image sharpness based on wavelet analysis given in the thesis is an effective and feasible method to improve sharpness of blurred noisy image.