Fast and Effective Interpolation Using Median Filter

Abstract

This paper proposes a fast and effective image interpolation algorithm using median filter. The interpolation algorithm is composed of two steps. First, a non-linear iterative procedure is utilized to interpolate the pixel whose direction can be easily determined by local information. Second, according to the introduced assumption that image interpolation can be regarded as a local image filtering process, the remaining pixels are interpolated by the proposed fast median filter method. Experimental results show that the proposed algorithm provides better performance than traditional techniques (e.g. bilinear interpolation, bicubic interpolation) both in subjective quality and objective quality with similar complexity. In particular, the proposed method is comparable to the well-known NEDI algorithm in visual quality, however, with much lower computational complexity. Therefore, the proposed algorithm can be exploited for real-time applications due to the merits of low computational complexity and good image quality.