论文部分内容阅读
基于超光谱图像的特点,提出了一种三维集合分裂嵌入式零块编码(3D SPEZBC)的超光谱图像压缩算法。该算法首先采用三维小波包变换有效地去除超光谱图像的空间和谱间相关性,然后对于所生成的每个二维子带利用基于集合分裂的方法进行零块编码,最后再采用基于上下文的自适应算术编码来进一步提高编码性能。实验结果表明,3D SPEZBC算法具有与三维嵌入式零块编码(3D EZBC)算法相同的压缩编码性能,在各比特率下编码性能均明显优于三维集合分裂嵌入式块编码(3D SPECK)、三维等级树集合分裂(3D SPIHT)和非对称三维等级树集合分裂(AT-3D SPIHT)算法,并且略好于多分量JPEG2000编码(JPEG2000-MC)算法。此外,3D SPEZBC编码算法不但可以提供较好的率失真性能,而且相对于3D EZBC编码算法可以节省大量的存储空间。
Based on the characteristics of hyperspectral images, a 3D hyperspectral image compression algorithm based on three-dimensional set split embedded zero block coding (3D SPEZBC) is proposed. The algorithm first uses three-dimensional wavelet packet transform to effectively remove the spatial and spectral correlation of hyperspectral images, and then uses zero-block coding based on set splitting for each generated two-dimensional sub-band, and finally uses a context-based Adaptive arithmetic coding to further improve coding performance. The experimental results show that the 3D SPEZBC algorithm has the same compression performance as the 3D EZBC algorithm, and the performance of the 3D SPEZBC algorithm is obviously better than that of the 3D SPECK, 3D SPIHT and AT-3D SPIHT algorithms, and slightly better than the JPEG2000-MC algorithm. In addition, the 3D SPEZBC encoding algorithm not only provides better rate-distortion performance, but also saves a lot of storage space compared to the 3D EZBC encoding algorithm.