Multi-View Observations Reconstruction with Modified Joint Sparse Representation

Sparse signal reconstruction can accurately reconstruct the original signal using a few atoms in a sparse dictionary, and it has been widely used in many fields such as radar, image, and speech processing. In the paper, we propose a modified joint sparse reconstruction method for sparse signal reconstruction under specific observation conditions. The proposed method is applicable to scenarios where multi-view observations are continuously obtained from different angles, such as continuous observations from multi-station radar. By applying different sparse constraints between inter station and intra station signals, the weak correlation of multi-station observations and the strong correlation of continuous observations within a single-station can be fully utilized to enhance reconstruction performance. The reconstruction experiment was conducted using multi-view radar observations based on the above scenario, and the results verified the effectiveness of the proposed method.