Ground Class Boundary Division Based On Tilt Photography Data Of Uav

Tilt photogrammetry is a cost-effective and flexible way to obtain high-quality and high-resolution geospatial data to depict the vertical and horizontal structures of the urban environment.In recent years,uav tilt photography technology has provided important technical support for cadastral mapping.Paper using unmanned aerial vehicle(uav)tilt photography collection data information,in the process of mapping using the ORB(Oriented FAST and Rotated BRIEF)operator image matching methods of extracting feature points,The edge on the basis of the generated map using global probability(Globalized aim-listed probability of a boundary,gPb)contour detection and simple linear iterative clustering(simple linear interative clustering,SLIC)super pixels of the image processing method,extracted from the map and deal with the boundary characteristics,elaborate class automatically generate boundary contour,which finished class cadastral control boundary contour work.The main research contents and work of this paper are as follows:1.Main components of uav tilt photogrammetry system,types of tilt photogrammetry and requirements for camera configuration are described.At the same time,the processing flow of this technology and the pretreatment processes of aerial photography such as image correction,image enhancement and denoising are analyzed.2.Analyze the influence of photo inclination on the accuracy of mapping,study and demonstrate the requirements of photo inclination with examples;The effects of optical distortion and aerial triangulation error on the accuracy of mapping are analyzed.The ORB algorithm in the image processing method is used to extract image feature points.In order to verify the effectiveness of the ORB algorithm,image matching experiments are carried out.An example is used to carry out a graph test,and the measurement accuracy is analyzed.3.For areas where the contours of ground objects are clearly visible and coincide with the boundary of ground objects,the demarcation method of the boundary of ground objects based on uav data is proposed.Based on global boundary probability(gPb)contour detection and simple linear iterative clustering(SLIC)superpixel,this method extracts and processes the boundary features of candidate sites from uav data and generates the boundary contour.The method is proved to be automatic and convenient for the boundary division of visible ground class by precision detection analysis.