Study On Urban Road And Vehicle Information Extraction From High Spatial Resolution Remote Sensing Image Based On Object Oriented:a Case Study

The rapid development of high resolution remote sensing image and commercialization makes it possible for research on urban traffic planning by using remote sensing data in large scale. Its wide application and abundant geographic information is greatly influencing the life of people. Remote sensing image has the unique advantages of continuous coverage on large spatial area, so the use of remote sensing image for unbern road information, especially network information and vehicle detection, is becoming a new research field.This paper centers on study on urban road and vehicle information extraction from high spatial resolution remote sensing image based on object oriented to solve the problems about gaining urban road information based on high resolution remote sensing image data by synthetically using the image processing theories, object-oriented and feature analysis. By using theory and software of traffic remote sensing and GIS, the paper systematically puts forward the process and method of urban road information extraction. First of all, it acquires high spatial resolution remote sensing image via "Google Earth" and "Daoge Google Map Interceptor"; secondly, it uses the object-oriented approach to compare and verify the two kinds of road network extraction by reasonablely selecting extraction method to conduct experiment on the extraction of transport hub and making the thematic map named "Middle of North Ring4Thematic Map"; finally, experiments are conducted, such as vehicle identification based on road network extraction, calculating the traffic flow space data (road space share etc.). It also analyzes the influence of the improvement of vehicle identification precision in two different levels of resolution images.Compared with manual work, the efficiency of experiment process and the accuracy of object extraction methods, have been improved more than existing methods. In this paper, the research provides the technical foundation for adhibition of high resolution remote sensing image in information acquisition for intelligent transportation system as well as traffic planning.