Image Filtering and Enhancement Based on Adaptive Triangular Meshes

In the procedure of image sampling, acquisition, coding and transmission, all images in different degrees may be polluted by noise, which lowers image quality. Image filtering aims at noise suppression and image enhancement sets itself a goal of improving visual effect by a series of techniques and enhancing image sharpness in order to transfer images into the forms that are more suitable for a person or a machine to analysis and processing. This paper proposes an optimal algorithm for image filtering and enhancement operations based on adaptive triangular meshes. Experimental results show that these operations directly working upon adaptive triangular meshes that represent gray-level images are significantly more efficient than their pixel-to-pixel counterparts. The execution upon triangular meshes is two orders of magnitude faster than the execution upon the original images.