The 3-D Reconstruction Of Urban Buildings Based On A Single High Resolution SAR Image

Synthetic aperture radar(SAR) is a kind of microwave remote sensing radar which can work in all-weather and all-time. Affected by speckle noise, it is very difficult to interpret SAR images visually. In recent years, with the appearance of high resolution SAR images, the extraction of buildings’ 3-d information has become more accurate. It provides the basic data support for the 3-d reconstruction of buildings. The technology of SAR provides a strong support for urban planning, detecting the all-time change of city and assessment of natural disasters.In this paper, we research the extraction of layover shaped as L on a single high resolution SAR image. The model and features of the layover are analyzed in great detail. Finally the 3-d reconstruction of buildings is done on the basis.Firstly, the principle of the SAR and the characteristic of SAR images is introduced, then the scattering characteristics of common buildings in SAR images is analyzed, and the principles and features of shadow,layover and secondary scattering is analyzed for two kind of common buildings, which provides a theoretical basis for the subsequent chapters. Secondly, the detection algorithms of the buildings is introduced, then we improve it according to the problem of precision and accuracy where median filter is used to enhance image quality, corrosion and expansion fill empty places and counting filter for filtering clutter. Finally, the CFAR algorithm is used to detect the outline of buildings, the Hough transform linear detection algorithm is used to extract the layover. After the 3-d information extracted, the 3d reconstruction of buildings will be finished.