Distributed compressed sensing is concerned with representing an ensemble of jointly sparse signals using as few linear measurements as possible. Joint reconstruction algorithm for distributed compressed perception was based on the idea of using one of the signals as side information, and then reconstruct other signals by the correlation between the side information and other signals. To resolve the complexity of reconstruction algorithms and reduce the measurements, two novel joint reconstruction algorithms for distributed compressed sensing based on joint sparse models were presented in this paper. Its application in signals and images processing was presented which are on the basis of demonstrating its feasibility. The result represent that the two novel joint reconstruction algorithms need fewer measurements for getting the same quality.