Algorithmic techniques for discrimination among similar targets

We investigate how imaging sensors can be exploited to support airborne radar persistent tracking of targets with similar kinematics. We study two image discrimination algorithms. One algorithm is based upon a chi-square test performed on the histograms of the images and the other is based on a two dimensional cross-correlation of the images. Both techniques may be applied to a variety of imaging sensors, including SAR/ISAR, EO/IR cameras, and HRR radar. A comparison of the two approaches is provided for the case of EO images with varying view angles and signal to noise ratios.