Application of Multi-Resolution Filter in Image Denoising

Many applications require the management of image data. A good approach should be efficient and detect noise in image of arbitrary shape. A new concept of Multi-resolution filter is proposed in this paper. It is suitable for images corrupted by speckled noise than any other filters. The implementation of a filter bank is expressed. Using multi-resolution property of wavelet transforms, we can effectively identify arbitrary speckled image at different degrees. We describe an application to speckle reduction in images, meanwhile we study and compare the performances of speckled reduction method with those that of the most often used adaptive filters, the result show that, with our method, the performance improve for efficiency.