Semi-automatic Extraction Of Remote Sensing Images In The Road Study

With the rapid development of the technology of remote sensing, we can get vast geographic dates. It's a very important question to automatically process and interpret these iconic data for us in this information era. As a main means to obtain the geometrical and physical information of the surface, information automatic interpretation of spatial images is an advanced research field in photogrammetry, remote sensing and computer vision research area. So semi-automatic road extraction is attracting more and more attention.In this paper, we focus on the semi-automatic road feature extraction from remote sensing images by means of human-machine cooperation. The main works are summarized as below:1. Introducing the background of research and the development trend of road extraction., and several main semi-automatic road extraction algorithms are listed and analyzed.2. Introducing the expressing and using of knowledge in automatic road extraction.3. Several algorithms for the preprocessing of road extraction and edge detection are experimented, and obtained some efficient algorithms, and several algorithms for the road edges, road center points, the width and the geometrical describes of road are discussed, the corresponding calculation algorithms are given then.4. Concept of fuzzy math is imported. Presents a robust and fast semi-automatic road extraction algorithms based on the theory of fuzzy operator to extract the road from low-resolution images..5. Using the approach of template matching to extract the road from high-resolution images. This approach utilizes a series of methods to enhance the ability of anti-noise. The results of experiment indicate that the road center curve extracted by this approach is fairy extract.