An Image Interpolation Algorithm Based on the Classification of the Pixels

Because the characters of image pixels near to edge and ones in the homogeneous area are different, we should use different interpolation algorithms to different pixels while interpolating. This paper introduces a new method for the image interpolation. First of all, it classifies image pixels as homogeneous and edge ones according to the threshold method, then it uses the lagrange interpolation algorithm to the homogeneous pixels, and the inverse distance weighted interpolation algorithm to the edge pixels according to the adjacent-similar principle. The experiment results show that the method in this paper gets better results than using only lagrange interpolation algorithm.