Research On Synthetic Aperture Radar Image Enhancement And Detection Technology

The technology of Synthetic Aperture Radar (SAR) has been developed in a high-speed in recent years for it抯 all-weather and high-resolution capabilities. Because of it抯 wide observation range, the development of computer science and modern signal processing, SAR has obtained day by day widespread application in many areas such as remote sensing, exploration of terrestrial resources, catastrophe forecast, military maneuvering and other civilian applications. Image processing is a quick developed new science, which has abundant contents. Pattern Recognition is to identify the same or the similar between an event and the specimen. We should combine the two technologies to gain more quick development in the direction of intelligence, and for more benefit to people all over the world. In accordance with the feature of radar data, we first use filter processing to reduce the noise. Many methods and technologies are researched for a special target such as line target due to the theory and technology of pattern recognition. The Contents is followed: a). At first, a theory research in the methods of image processing and pattern recognition is given. Some methods, such as histogram balance, are given a necessary complement in accordance with the feature of radar data. Some experiments were proceeded using the data available. b). Due to the existence of speckle noise in radar image, some base methods of speckle noise attenuation are researched to detect target interested. Thorough research of weighted median filter is proceeded and gained some valuable results, which provide a base of the detection work next. c). In accordance with some special targets, many methods of line detection are researched using the available data. Finally discussions about coastline and road detection are given.