Recently one of the most important problems in image restoration is to remove the additive noise produced by the imaging system without blurring the fine details of the image. This problem arises in machine vision, remote sensing, computer tomography, and other imaging systems.
During the past decades, much effort has been devoted to the development of detail-preserving filters. However, the outcome is still limited by the quality of pictures.
This paper proposes detail-preserving multilevel KAVE filters. It is an improved classical KAVE filter based on the concepts of multilevel operation. The characteristics of this filters are analyzed and compared with several other known filters by their ability to retain fine details. Simulation studies show that proposed filters preserve fine details of the images very well but still have limits in some kinds of images.
The experimental results help researchers in improving the detail-preserving filters on the base of multilevel operation.