Research On Theory And Methods Of Remote Sensing Video Imaging Based On Compressed Sensing

The contradictions between massive remote sensing data acquisition,large bandwidth and limited sampling and storage space on satellites could be resolved by applying the compressed sensing(CS)theory to remote sensing video imaging.Around the CS-based remote sensing video imaging system,the remote sensing video imaging application of compressed sensing theory was researched in this paper from three aspects including the modeling of CS-based video imaging system,reconstruction models & reconstruction algorithms design,and sparse representation dictionaries design.According to the video imaging model,the distributed compressed video sensing(DCVS)coding models are set up,and the DCVS-based decoding algorithms are designed,which is the foundation of video reconstruction and sparse representation dictionaries design.The residual distributed video reconstruction model based on bilateral motion estimation and algorithms applying on the model are designed,on the basis of bilateral motion estimation.In allusion to limited communication bandwidth in practice,an adaptive sampling and multi-hypothesis residual reconstruction model,in which the measurement ratios is controllable,and algorithms are proposed.What's more,the models and algorithms are followed by numeric simulations to check their effectiveness.The block-PCA dictionary and residue learning dictionary orienting residual reconstruction algorithms are proposed,under the DCVS frame.Based on Householder transform theory,a multi-hypothesis Householder dictionary is designed.As well as the former chapters,the numeric simulation is executed to demonstrate the performance of various dictionaries.