New enhanced image quality interpolation algorithm based on edge detection
Zhao Yong · Jisuanji gongcheng yu sheji · 2007
The proposed algorithm first detects edges in the high-resolution image obtained by classical interpolation algorithm.Then the edge's neighbors are amended by a self-relativity interpolation algorithm based on multi-direction gradient variety.And small ratio cycle amendatory method is used in multi-time enlargement.Experimental results show that the algorithm can effectively avoid blurring and artifacts edges and generate high quality images with less data.If this image is RGB,color fusion is processed first.Then image edges are detected.And the visual quality of the color images more prominent.