中文摘要：

根据超光谱图像特点和应用环境要求,提出一种基于DCT变换域分布式信源编码的超光谱图像压缩算法.编码端首先进行线性预测,然后按DCT子带顺序进行比特平面编码,解码端使用关键帧重建边信息进行LDPC解码.在编码端使用部分像素点进行线性预测,从而减少了编码端的运算量.在解码端,利用已解码子带信息进行基于块的迭代线性预测,使用优化后的边信息解码后面的子带.与传统算法相比,本算法在编码端的运算量更少,需要的存储空间更小,满足超光谱图像压缩系统要求,易于硬件实现.

英文摘要：

Based on the analysis of the hyper-spectral images,a new compression algorithm based on the DCT transform domain distributed source coding is proposed.Our algorithm performs the bitplane encoding at the encoder with the DCT subbands order,while using the key frame to reconstruct the side information for LDPC decoding at the decoder.Few pixels are adopted to perform linear prediction at the encoder,thus reducing the complexity.Subbands previously decoded are utilized for iterative linear prediction based on blocks at the decoder,and following subbands are decoded with optimized side information.Compared with conventional algorithms,the proposed algorithm efficiently reduces the cost of computation and memory usage at the encoder,which facilitates the hardware implementation.