Recursive Least Squares for Near-Lossless Hyperspectral Data Compression

The hyperspectral image compression scheme is a trade-off between the limited hardware resources of on-board platform and ever-growing resolution optical instruments. Predictive coding attracts researchers due to its low computational complexity moderate memory requirements. We propose near-lossless prediction-based that removes spatial spectral redundant information, thereby significantly reducing size images. This predicts target pixel’s value via linear combination previous pixels. weight matrix predictor iteratively updated using recursive least squares filter with loop quantizer. optimal number bands for prediction was analyzed experimentally. results indicate proposed outperforms state-of-the-art methods in terms ratio quality retrieval.