2D spatial model matching using HRR multi-radar data

In this paper, the problem of target classification from multiple high range resolution (HRR) radars data is studied. The use of multi-sensor angle-diverse data aims at shortening of the required time before a decision is made, as compared to using single-sensor data. In order to avoid the high-dimensional HRR profile databases, involved in the classic automatic target recognition (ATR) approaches, the classification is based on 2D spatial target profile matching. Stochastic target modeling is further applied, as a more robust way to deal with more realistic target scattering behaviors at variant aspect angles. Raw video measurements are used, with the objective to circumvent problems due to missed detections and false alarms in the thresholded profiles. This paper proposes a new method, where the above elements are combined. Some first encouraging simulation results are given.