Super resolution imaging is a kind of data adaptive imaging technology, which can be applied to the group target recognition of ISAR imaging. At present, the error caused by the optimal weighting vector can cause the performance of the beam forming device to drop dramatically. However, the imaging effect of the super resolution imaging algorithm is greatly improved after the addition of the quadratically constraints and subspace constraints, and has a good effect on the target recognition.Based on the Capon as the starting point, this paper analyzes the advantages of the two constraints and subspace constraints of the super resolution imaging algorithm, and compares the diagonal loading algorithm and the beam forming algorithm based on the feature space, and finally shows the results through simulation.