Anomaly detection based on an iterative local statistics approach

We introduce an iterative anomaly detection algorithm. The algorithm is based on an iterative characterization of the clutter in a feature space of principal components, and a single hypothesis scheme for the detection of anomalous pixels. The iterative procedure gradually reduces the false alarm rate while maintaining a high probability of detection. Morphological operators are subsequently employed for extracting the sizes and shapes of anomalous clusters in the image domain, and identifying potential targets. Experimental results demonstrate the robustness of the proposed approach with application to sea-mine detection in sonar imagery.