A novel approach for the automatic detection of punctual isolated targets in a noisy background in SAR imagery

Several distinctive capabilities of satellite-based synthetic aperture radar (SAR) systems make them a powerful tool for Earth observation purposes. In fact, SAR systems can provide an efficient, quick and cost effective coverage of any area independently from weather effects and from the day-night cycle. Nevertheless, the automatic interpretation of SAR images usually constitutes a complicated task due to the presence of speckle and to the reduced dimensions of the target compared to the sensor resolution. This papers aims at presenting a novel algorithm for the detection of punctual isolated targets in a noisy background based on the wavelet theory. It will be shown that this can be useful for example for ship detection purposes