Spatial–Temporal Expanding-Accumulating Strategy-Based BACF Algorithm for ViSAR Dim and Small Target Tracking

A large number of publications on detection and tracking have appeared in the field of video synthetic aperture radar (ViSAR) due to its capability of forming high-resolution videos for a region of interest. In this work, a novel spatial-temporal expanding-accumulating strategy based background-aware correlation filter (BACF) algorithm, namely STEA-BACF, is proposed to tackle theViSAR tracking problem. In the proposed STEA-BACF algorithm, multiple response values are selected from the response map to generate multiple candidate bounding boxes. Each candidate bounding box is then expanded in both spatial and temporal domains to obtain a number of sub-candidate bounding boxes, resulting in a tree-structure composed of multiple candidate bounding boxes, the response values of multiple candidate bounding boxes are accumulated over multiple frames. Finally, a backtracking algorithm is used to select a path with the maximum accumulated response values, thus determining the target as well as its trajectory simultaneously. A number of experiments are conducted using the ViSAR data released by Sandia National Laboratory. The experimental results show that the proposed STEA-BACF algorithm outperforms several state-of-the-art model-based trackers.