Fast detecting and locating groups of targets in high-resolution SAR images

A new algorithm for detecting and locating groups of targets in high-resolution SAR images is presented. Firstly, a global constant false alarm rate (CFAR) detector is utilized to locate the potential target regions. Then, the size, shape, and contrast features of each region are computed for target discrimination based on voting decision. As targets are often deployed in well-defined groups and all targets within one group have a certain relationship in their positions, it is possible to extract the whole target group and diminish clutter false alarms. The experiment results show that the proposed algorithm significantly reduces the regions revisited by an automatic target recognition (ATR) system, and false alarms can be greatly diminished.