Multi-grid edge models for magnifying digital images

Abstract

A set of discrete ideal step edge models have been designed for small windows of pixels. These models consist of pairs of low and high resolution ideal step edge patterns. Based on the assumptions that within a small image region, the underlying edge structure and the average pixel intensity of the low and high resolution image samples should be identical, a small window of low resolution pixels are mapped to a high resolution lattice. The method involves mainly table look-up operation and is therefore computationally efficient. The models have been applied to the reconstruction of high resolution images. Simulation results show that the images expanded by the new method have sharper edges and lower reconstruction errors than those produced by traditional nearest neighbor, bilinear and bicubic interpolation methods.