Simulation Experiment of DPCM Compression System for High Resolution Multi-spectral Remote Sensing Images
Wei Du,Yi Wu,Mei Tian,Wei Hu,Zhidong Liu
DOI: https://doi.org/10.1007/978-3-031-38651-0_30
2023-01-01
Abstract:Compared with natural images, remote sensing images are rich in texture information and have poor spatial correlation, so it is difficult to obtain high compression ratio only by removing spatial redundancy in the spectrum. In order not to lose the original information obtained by detectors, remote sensing data usually require lossless compression. In this paper, the simulation experiment of DPCM compression system for high resolution (HR) multispectral remote sensing image is carried out, and a new algorithm for HR multispectral remote sensing image compression using wavelet transformation (WT) and piecewise DPCM hybrid coding is introduced. Firstly, the algorithm uses DPCM method to remove the inter-spectral redundancy of HR multi-spectral remote sensing images, and then uses WT to encode the residual image after removing the inter-spectral redundancy. This algorithm can not only realize near lossless compression of HR multispectral remote sensing images, but also realize lossy compression with different accuracy conveniently. The simulation results show that the entropy value of the image is reduced after the compression of this algorithm. The algorithm in this paper solves the problems existing in calculating the basis vector by covariance estimation and feature decomposition, and it only needs a few bits, but the calculation is very effective. It shows that the algorithm in this paper can achieve ideal compression effect.