Extraction Of City Road From High Resolution Sensing Images Based On Hough Transform

High-resolution remote sensing images for city reflect the various details of urban space. Town road as a skeleton, its recognition and extraction, information update plays an important role on urban dynamics, terrain analysis, urban construction. Extracting road from high-resolution remote sensing images is an important sector of urban dynamics, digital city and urban construction. Accurately and efficiently extract road information from high-resolution remote sensing images is a hot research for scholars in recent years. At present, the results of most of the road based on high-resolution remote sensing images belongs to the semi-automatic extraction of road extraction, with the limitations of large and small scope. Further study of road extraction from high resolution images has more theoretical and practical significance with the rapid development of the city as well as the continuous improvement of the resolution remote sensing images.Hough transform has good Robust character used in extracting rules straight and curves in Digital Image Processing, the used of Hough transform in road extraction from high-resolution remote sensing images still valid, but primarily for straight road extraction, the extraction of curve road mainly used to extract is scarcely at present, and Hough transform need large parameter space, computing speed slowly, generating restricted roads for extraction from high-resolution remote sensing images. Paper discusses the road extraction from high resolution remote sensing image based on Hough Transform, dissert the basic theory of image filtering and edge detection, summarized description of the advantages and disadvantages of those approach, introduces the principle of Hough transform algorithm and the computational process. Paper improves the traditional Hough transform in road extraction for the characteristics and shortcomings, Optimized Hough transform algorithm based on connected domain of image segmentation technology, extracting the curved road and straight road based on Hough transform, experiment results show that Optimized Hough transform algorithm based on connected domain of image segmentation technology can improve the computing speed, improve road extraction consequent.