Object tracking - an art based phase filtering approach

Finding and tracking a special pattern or template in a series of images with high frequency background, such as terrain image generated by satellite, is an important topic. A pattern in a satellite image has not the scaling problem but has the rotation problem since the satellite gets these images at a fix altitude. The phase only match filter (POMF) method is a kind of method that can recognize a given pattern precisely even that pattern has been rotated but not been scaled. However, the recognizing performance of the method is affected by the high frequency noise and size of image. In order to over come these problem, a preprocess stage is proposed to remove most background and reduce the size of image through artificial resonance theory (ART) neural network. Moreover, a mechanism that combines the preprocess stage and POMF method to register and track a pattern in a series of satellite images is proposed as well. The proposed algorithm has been examined by means of several satellite images. According to the experimental results, the proposed method demonstrates that the performance of POMF can be successfully improved by reducing the high frequent noise of the background in the satellite images.