Open-set HRRP recognition method based on joint sparse representation

Abstract Aiming at the problem of multi-view high-resolution range profile (HRRP) target recognition under open set conditions, we proposed an method based on joint sparse representation (JSR), which solves low rate traditional methods conditions. This is applied to background radar single-station observation. JSR used solve reconstruction error HRRP by over-complete dictionary, while extreme value theory (EVT) model tailing matching and non-matching categories transform identification into a hypothesis testing problem. During recognition, use determine candidate class, scores class are obtained according confidence tail distribution, weighted sum two as category criterion finally or outside library. can effectively relevant information between HRRPs improve performance The algorithm tested with data generated from MSTAR inversion, results show that better than mainstream method.