Region of interest (ROI) determination is a first and crucial step performed in an automatic targetrecognition (ATR) system. The goal of ROI determination is to identify candidate regions that mayhave potential targets. To be most effective, this initial detection (or focus of attention) stage mustreject clutter (noise or countermeasures that provide target like characteristics), while ensuring thatregions with true targets are not missed. We present a novel approach to ROI determination insynthetic aperture radar (SAR) images for ATR based on the premise that regions with targets wouldrequire a model with more free parameters to smoothly approximate the magnitude of the return.Toward that end, we use a sigmoidal multilayered feed-forward neural network with selected lateralconnections between hidden layer neurons to approximate the return in disjoint square patches of theSAR image. This network probably uses as few neurons as possible to produce a desiredapproximation and thus enables the determination of the number of parameters used in approximatingthe return in an image patch. Those squares of the image that require a large number of neurons(more free parameters) are then labeled as ROIs. Results obtained with synthetic and real-world SARimages are used to demonstrate the effectiveness of the proposed method. A significant advantage ofthe proposed method is that it does not require the presence of a training data set, which, given thevariability in SAR images and target signatures, is difficult to obtain.
展开▼
