In this paper, a compressed sensing signal reconstruction algorithm that based on Bayesian compression perception theory is proposed. It can be applied to signal compression and transmission as well as reconstruction. The new algorithm inverted the compressed sensing problem into a linear regression problem. Firstly, then deduced an iterative relationship of the resulting vectors gradually, at last got the exact reconstruction of the original signal by iteration. The simulation experiment exploted that the Bayesian compressive sensing algorithm have a good reconstruction effect used in one-dimensional and two-dimensional signal processing and the reconstruction.