A new method for ISAR target detection based on chief axis analysis and block feature extraction

The traditional methods for feature extraction of ISAR(Inverse Synthetic Aperture Radar)target images suffer from the following two shortcomings: (1) The energy distribution of the target is not considered; (2) The influence of the weak scattering points in describing target details is ignored. We present a new method for ISAR target detection to overcome these shortcomings. First, the target chief axis direction and the target centre are obtained by the strong scattering point coordinate matrix. Then, along the main distribution range of the target energy, block segmentation is carried out on the target. The extracted blocks prevent the strong scattering points from restraining the behavior ability of the weak scattering points' geometry structure. Therefore, the target image obtained by our method can reflect not only the visual characteristic of the target, but also the detail of the energy distribution.