Research On Key Technologies And Applications Of 1:500 Large-scale UAV Mapping

In recent years,the rapid development of aero photogrammetry technology and the gradual improvement of aerial photogrammetry have made the accuracy of aero photogrammetry improved continuously.The high precision and high resolution remote sensing images are gradually becoming normal by aerial photogrammetry.As an important means to quickly and efficiently obtain and update the geographic information data,aero photogrammetry has unique advantages in many fields,such as disaster emergency treatment,geographical condition monitoring and so on,and is gradually becoming an important subject of national infrastructure construction.With the development of UAV technology,the unmanned aerial photogrammetry system by flight platform equipped with digital camera image sensor in UAV,with GPS and IMU technology to obtain the ground image data.Through data processing,we can get geospatial information and their relationship.Compared with the traditional aerial photogrammetry,the UAV aerial photogrammetry system has the advantages of high resolution,high efficiency,flexible maneuver and low cost,and is widely used in small range and large-scale mapping.Taking the low altitude photogrammetry of unmanned aerial vehicle as an example,this paper introduces the key technology and work flow of low altitude photogrammetry in detail,and verifies the precision of the large scale mapping of unmanned aerial vehicle through the engineering test,and realizes the automatic extraction of the building based on the digital orthophoto map and the digital surface model obtained by the low altitude photogrammetry.The main contents are as follows:(1)Three methods of large scale mapping using unmanned aerial vehicle(UAV),including UAV low altitude photogrammetry,UAV laser radar and tilt photogrammetry,are introduced.The working principles and advantages and disadvantages of the three methods are compared and analyzed.(2)The key technology of UAV low altitude photogrammetry system is studied.The non-measurement camera is calibrated by direct linear transformation,and the calibration parameters are used to correct the distortion of the image.A POS assisted aerial triangulation with exposure delay is proposed by analyzing the shortcomings of the traditional aerial triangulation.The advantages and disadvantages of the SIFT feature matching algorithm and the improved PCA-SIFT algorithm are verified by the specific experiments.(3)The work flow of the UAV low altitude photogrammetry system is introduced in detail,and the digital orthophoto map and digital surface model of the target area are obtained by data processing,and the processing results are checked with the measured data.It is proved that the unmanned aerial vehicle(UAV)aerial survey can reach the requirement of the 1:500 scale mapping accuracy.Compared with the result of adding sparse control points,it is proved that adding control points can significantly improve aerial survey elevation accuracy.(4)In combination with the UAV Orthophoto Image and DSM data,the DSM image is two valued by selecting the elevation threshold and using the SVM classifier to classify the unmanned aerial vehicle(UAV)orthophoto images to get the vegetation coverage area,and the DSM data that only contains the building information is superimposed with the DSM,and the building area is extracted by the small area removal and morphological operation.Finally,the building boundary is determined by Grabcut algorithm,and the results of the building extraction are evaluated by three evaluation factors,positive rate,integrity rate and detection quality.