The Research Of Remote Sensing Image Vehicle Information Extraction Based On Machine Learning

With the economic development of our country, the construction of the urbanroad traffic has become the top priority.And the increase of car ownership has becomeone of the important factors for hindering the development of road traffic. How toreform the urban road network, so as to guarantee the safe and efficient operation ofurban traffic, has gradually become the focus of research. Therefore, relevantdepartments need to adopt a method that can collect traffic information which iswide range, high accuracy and practical effect. In order to obtain more accurate andcomprehensive information of the related region to be analyzed, and solve thedevelopment of urban traffic problems.For the above problems, this paper carries out the high resolution remote sensingimage of research on the algorithm of vehicle information extraction technology. Bythe characteristics analysis of vehicle targets in high-resolution remote sensing image,establish vehicle sample and non-vehicle sample, and form the sample library. On thebasis of the original image, after image preprocessing, training the support vectormachine(SVM) learning with the SURF feature which was extracted in the image.And then, according to the effect of the test sample classification, improve theparameters of SVM to achieve the optimal classification results. In full search, amethod of half feature sliding window search is put forward. Setting the classifierthreshold to improve the accuracy of target search and reduce the amount ofcalculation. At the same time, in order to optimize the algorithm, we add the roads’mask file, make the data more reliability.Using high-resolution remote sensing image to extract the vehicle informationinside the city’s road, not only improves the degree of China’s transportation industryinformatization and the road network analysis information source, but also providesan important decision-making information for the development of the intelligenttransportation. And prompts the transportation industry in China enters a new level.