I want to enhance an image which has been corrupted with speckle noise by using a dictionary technique, the input image should be of size 256x256 or 512x512. the output image should be distinct from the input image that is to say the output should be the enhanced version of the input image. Again the input image should be an image which has been corrupted with speckle or an image that we can add speckle noise to it in different levels ( different variance of noise). The following Papers can be of help for you to understand what am trying to say.
1. Image detail enhancement using a dictionary technique by A. Choudhury, P. van Beek and A. Segall.
2. Dictionary learning based image enhancement for rarity detection by Weifeng Liu, Xiaomeng Wang and Yanjiang Wang.
3. A HYBRID WAVELET-SPATIAL DENOISING FILTER by Adib Akl, Charles Yaacoub