In this paper, we propose a new algorithm named Optimal Recursive Bidirection Prediction (ORBP) for HyperSpetral Image (HSI) compression. Recursive Bidirection Prediction (RBP) is a good algorithm for HSI inter-band prediction. It's simple, efficient. But the algorithm can't get optimal prediction result under the sense of SNR. In this paper, the linear model of HSI is established, and the best prediction is deduced under the sense of SNR. The proposed method can get lower entropy after prediction. Computer simulation results show that compared with the traditional algorithm RBP, the proposed method ORBP has a great improvement by about 9.2738dB in SNR in average.
展开▼
