Research On Extraction Of Roads And Rivers From High-resolution SAR Image

Synthetic aperture radar is an active radar, which is capable of providing of all-time and all-weather detailed ground mapping data and images. Furthermore, SAR is highly penetrable under certain bands that could go through vegetation and shallow soil. Therefore, it is widely used in geology, agriculture, forestry, hydrology, environmental monitoring, national defense and other fields. SAR imaging technology has been developing rapidly in recent years, which has reached sub-meter resolution, and the quality of the SAR image has been signifi cantly improved. Roads and rivers, as an very important basic geographic information, holding a great significance for disaster prevention, urban planning, digital cartography and so on. Therefore, roads and rivers extraction become a hot topic in the field of remote sensing applications. This thesis proposed two ways for the target detection of the roads and rivers based on the characteristics of the high-resolution SAR image data.1. This thesis proposes a new algorithm for road extraction in high resolution SAR image, which combines parametric kernel gragh cuts and mathematical morphology. First of all, the roads in the SAR image are initially segmented using parametric kernel gragh cuts, and void spots are filled and road edge is smoothed by mathematical morphology; secondly, road extraction is obtained by the degree of matrix, the improved length-width ratio, complexity and other factors to remove the false alarm based on the geometrical characteristics of the road. Finally, road centerline is extracted using mathematical morphology, and the broken road is connected by the segment proximity and directional consistency, and then the road width is restored by mathematical morphology so now a perfect road extraction of high-resolution SAR images is achieved. The results showed that the algorithm can effectively suppress speckle noise and extract the road accurately and completely without performing SAR image preprocessing.2. This thesis proposes a new algorithm for river extraction in high resolution SAR image, which provides two extraction algorithms: the wide rivers extraction and the narrow river extraction. The wide rivers extraction algorithm is to use the level set segmentation as the first step based on the distribution of G~0, then, to obtain the wide river extraction by the degree of matrix remove the false alarm based on the geometrical characteristics of the river. The small river extraction algorithm is to adopt the histogram equalization to enhance the river features in SAR image, and then use the improved lee filter to suppress speckle noise, completing the SAR image preprocessing. Next step is identical with the width river algorithm. Experimental results showed that the false alarm rate and missing alarm rates are very low, and the most of the rivers goal can be detected. This algorithm could meet the requirements of high-precision river extraction.