Remote Sensing Images Of Road Extraction Method

With the rapid development of remote sensing technology, remote sensing data processing and gain useful information is a hot topic for many scholars.In many areas of remote sensing, road extraction and segmentation is an important technology, with good practical and military value. It has widespread applications,so it is an important subject with practical value and research value.Based on the related field of remote sensing technology, Learning and summarize the advanced experience of road extraction, establish the improved mathematical models for road with the common image processing technologies.We learn the strengths and weaknesses of RGB color space and HIS color space, establish a mixed color space model to filter the invalidity pixel of remote sensing image and final achieve the framework extraction of the road.Based on the geometry characteristics of roads in remote sensing images, propose an regional expansion methods based on quadrilateral, to extract the road, the system is developed in the VC6.0 platform. It achieves road extraction from remote sensing image,and meeting the basic color image processing and operation.Experiments show that this method is an effective extraction method.