A New Fractal Hyperspectral Image Compression Algorithm
Yun Chen, Ruidong Gao · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2015
Because of the great data of the hyperspectral image, it is not good for storage and transmission.To work out a high efficiency compression method for hyperspectral image is essential.Prediction encoding is easy to realize and has been studied widely in the hyperspectral image compression field.High compression ratio, resolution independence and fast decoding speed are the main advantages of fractal coding which makes full use of the local self-similarity existing in images and is considered as a promising compression method.However, the application of fractal coding in the hyperspectral image compression field is not widespread.In this paper, we propose a new fractal hyperspectral image compression algorithm.Considering the noises which exist in the hyperspectral image and the integrity of decompression, Firstly, intra-band prediction is implemented to the first band.Checking the noise whether the current encoding band is greater than the threshold value.If it is greater, the current encoding band will be considered to be a noise band and perform intra-band prediction.The first non-noise band which precisely followed the noise band performs the intraband prediction encoding.The rest of the bands will be encoded by modified fractal coding algorithm.The proposed algorithm can effectively exploit the temporal-spatial correlation in hyperspectral image, since each range block is approximated by the domain block which is of the same size as the range block in the adjacent band.Experimental results indicate that the proposed algorithm provides very promising performance at low bitrate.