Research On Compression Algorithm For Time Modulated Imaging Spectrometer Data

Interference hyperspectral image is remote sensing image. It is widely applied in meteorology, environmental monitoring and military and other fields, and it has high spectral resolution with large amount of data. The research in it becomes one of the key problems of the application of its success. Therefore compression algorithm of Interference hyperspectral image has become an urgent problem to be solved in spectrum technology field. According to the characteristics of time modulated interference hyperspectral image, this paper researches on the compression algorithm of interference hyperspectral image.For the practical application of hyperspectral images in aerospace field, this paper made a deep research on interference hyperspectral image. It is analyzed from three aspects. Firstly, according to band pass characteristics of interference hyperspectral images, the image is decimated in order to reduce data size. Secondly, by analyzing the interference characteristics of hyperspectral images, combined with the characteristics of SPIHT, it puts forwards a SPIHT compression algorithm based on sub-band mapping transform. According to the similarity of interference hyperspectral image, a compression algorithm based on interference waveform sequencing is proposed.(1) A band-pass decimator is designed for interference signal. It is used to decimate the interference hyperspectral image. In order to improve the efficiency, decimator is designed with poly-phase structure in the paper. With this structure, the original filter is decomposed into multiple sub-filters. In this way, the speed of decimator is improved.(2) An algorithm of SPIHT based on sub-band map is proposed. The method combines the characteristics of interference images and SPIHT coding algorithm. It uses the sub-band mapping method instead of wavelet compression part in SPIHT. After the sub-band mapping transform, large values is mapped to "LF" area and small numerical value is mapped to "HF" area. The distribution of remap image is similar to the distribution of image with wavelet transform. Without wavelet transform, the method not only improves the coding efficiency, but also improves the compression ratio. So the method is suitable for the application of satellite.(3) A compression algorithm based on sort prediction is proposed. Because interference signals are similar, image redundancy can be effectively removed with the linear prediction. With similarity sequence, the similarity of image can be further improved. So the prediction error is reduced and compression performance is improved. Finally, it encodes the prediction error with SPIHT based on sub-band map to further improve the compression performance.