Block-based compressed sensing algorithm for image compressed and transmission in visible spectral remote sensing imaging system

High-resolution visible spectral remote sensing images enriches human cognition of the world, also brings challenges to the image storage and transmission. Traditional compressed methods can reduce the transmission bandwidth, however at the cost of long computing time, high algorithm complexity, and extraordinary errors. Compressed sensing can reconstruct high-quality images which are almost the same as the original images. It provides a new method for remote sensing image compressed and reconstruction. This paper analyzes the advantages of block-based compressed sensing (BCS) algorithm, and proves that BCS algorithm is superior to non-BCS algorithm from the perspective of algorithm complexity. This paper modifies the smoothed projected Landweber (SPL) algorithm to make it suitable for the visible spectral image. According to the simulation results, it is obvious that the BCS algorithm can effectively reduce the transmission cost and ensure the reconstruction quality of the compressed images.