Image Enhancement Based On Wavelet Multiresolution Analysis And Its Application

An image system consists of acquisition, transmission, reception and display, every part of which can be disturbed, so that the image equality can be degraded. Elementary requirement of image enhancement is how to deal with these degraded images to meet our need. One of main contents of image processing is image enhancement. Now, there are many methods of image enhancement, but they will more or less bring bad effects when enhancing images. Wavelet analysis has predominant advantages in image de-noising and image enhancement, due to its extracting multi-resolution characters of signal and making the difference between noise and signal very clear. Image de-noising and image enhancement based on wavelet multi-resolution are studied in the thesis. The main works are as follows:Firstly, we compare image enhancement in wavelet domain with that based on unsharp masking and indicate that there is intrinsic relation between them, namely they both enhance images by enhancing the high frequency coefficients. Because after an image is decomposed in wavelet domain, every high frequency can be enhanced. However, if unshrap masking is used, only one high frequency channel is enhanced. So unsharp masking is a special case of image enhancement based on wavelet multi-resolution.Secondly, wavelet de-noising and edge detection are merged into image enhancement, that is, the image de-noising and enhancement are accomplished at the same time. After an image is decomposed, its edges, details and noise will exist in high frequency. Noise can be removed by wavelet de-noising, and edges and details can be enhanced in high frequency, hence the goal of enhancing and de-noising at the same time can be attained. The de-noising and enhancement are linked by enhancement function. An image can be de-noised and enhanced by adjusting the parameters of enhancement function.Thirdly, the enhancement functions are studied in detail. There are two classes of the enhancement functions: linear and nonlinear. Their merits and shortcomings are analyzed. A new method to choose the parameters of these functions is brought forward.Finally, we have studied how to select wavelet bases. Several typical wavelet bases are adopted, and the simulation results are analyzed. Some advices selecting wavelet bases are put forward. The results are compared with those of other methods, too.