(An) adaptive image Iiterpolation scheme based on local gradient features

Abstract

In this dissertation work, image interpolation using an inverse gradient is studied. Two adaptive interpolation methods based on applying an inverse gradient to conventional bilinear and bicubic interpolation are proposed. In simulations, the proposed methods exhibited a better performance than conventional bilinear and bicubic methods, particularly in the edge regions. In addition, the proposed methods can be used irrespective of the magnification factor (MF) and easily implemented due to their simple structure. Based on using mean of gradients, a globally adaptive technique to automatically determine sharpness constant necessary for the proposed adaptive methods is proposed. This globally adaptive technique for sharpness constant shows negligible performance degradation compared with the optimum value of sharpness constant obtained by heuristic approach, but still the proposed adaptive methods using this globally adaptive technique for sharpness constant show superior performance compared with those of the conventional methods. Furthermore, an efficient structure of inverse gradient kernel is proposed and shows nearly equivalent performance compared with the original inverse gradient kernel. The complexity of adaptive bicubic interpolation using this efficient inverse gradient kernel is approximately twice that of the original bicubic interpolation.