Acceleration of fractal video compression using FFT

Abstract

Fractal based video compression promises to be an attractive technique due to its resolution independence and high compression property. The fractal video coding takes an advantage of self similarity within a frame and also between the successive frames in a video for compression. A new full search motion estimation algorithm is presented in this paper that uses FFT based circular cross-correlation as a full search block matching criteria. The processing speed of proposed block matching algorithm is higher in comparison with conventional full search block matching algorithm. This is done due to the elimination of repeated search in fractal encoding process for the best matching block in a large domain pool. Experimental results show that the proposed algorithm achieves higher processing speed as compared to conventional full search fractal encoder with equivalent quality of decompressed video for the same level of compression ratio.