Terrain Classification Based On Multi_scale Texture Analysis Using SAR Image

Synthetic Aperture Radar (SAR) san observe all_ weather , all day and penetrate some object .Therefore , it is widely applicatied in many areas, such as natural disaster monitoring, resource survey, military objective reconnaissance , change detection and so on. Nowadays SAR is one of the most active field in radar and remote sensing. But how to recognize object and acquire information from SAR imagery is a hot point of research.Because of the acute speckle of the microwave-imaging machine, especially for the single frequency, single look and single polarization SAR images, it is very difficult to extract image characteristics in the image pixel level. Therefore , it is a puzzle to study the classification modal, increase the class number and improve the detection ratio according to the classic classification algorithms.Texture analysis is an important tool in pattern recognition and image processing. The analysis of texture in an image provides an important cue to the recognition of objects. It will improve the understanding of images greatly.GrayJLevel Co-occurrence Matrix (GLCM) is proved a good method of texture analysis. In this paper, the GLCM is analyzed from two aspects which contain the directions and the scales of the GLCM. The result of the texture analysis shows that the textural features satisfy the SAR image classification requirements.