Aerial Image Mosaic Of Low-altitude Fixed-wing UAV Based On POS Data

With the popularization of UAV technology,ground image acquisition by UAV has been applied in many fields,such as rescue and disaster relief,topographic survey,forestry management,agricultural disaster monitoring,marine pollution monitoring,transmission line patrol and so on.For aerial photography of low altitude fixed-wing UAV,each image can not cover the whole survey area separately.To obtain the overall information of the survey area,it is necessary to splice the single image into a panoramic image covering the whole survey area.Scholars at home and abroad have done a lot of research on aerial image mosaic of low-altitude fixed-wing UAV.According to the characteristics of low-altitude aerial image,many image mosaic algorithms are proposed.Some of these algorithms focus on mosaic speed,some focus on mosaic quality,and often ignore the large amount of aerial image mosaic in the computer.The requirement of storage makes these methods have higher requirements for computer hardware.In order to overcome the problem of small memory and ensure the quality of mosaic when personal computers mosaic large quantities of Low Altitude Aerial images,this paper mainly studies from the following aspects:1.Aerial image pretreatment for low altitude fixed wing UAV.In order to improve the quality of panoramic image after stitching,aerial image pretreatment technology is needed to deal with the problems existing in the original aerial image,including correcting the image point displacement caused by objective lens distortion with distortion correction parameters,and eliminating the unfavorable factors such as haze and weak illumination intensity with image enhancement method of histogram equalization.The brightness of the image is improved by the influence of the image.The image is smoothed by the median filter of the 5 × 5 scale filter to remove the noise in the image.2.Research on aerial image mosaic of low altitude fixed wing UAV based on POS data.The aerial image of low-altitude fixed-wing UAV has tilt error due to the influence of the attitude of the flying platform,which needs to be corrected geometrically.(1)In order to improve the accuracy of the geographic coordinates of aerial images based on the classical least squares method,this paper uses the global least squares method to fit the coordinate data of the principal point of the object.The global least squares method takes the error of the coefficients into account,so the accuracy of the fitting results is more reasonable and the accuracy is improved.(2)In order to overcome the problem of small memory of personal computer in large-scale aerial image mosaic,this paper calls the function of image mosaic data set in ArcGIS Engine,which can realize aerial image mosaic in batches by establishing image index in spatial database,and effectively overcome the problem of small memory of computer.Experiments show that the low-altitude fixed-wing UAV mosaic method can achieve a large number of aerial images mosaic,greatly reducing the computer memory requirements.3.Design and implementation of low altitude fixed wing UAV aerial image mosaic system.In order to realize the function of UAV aerial image mosaic in the same system,a low altitude fixed UAV aerial image mosaic system is developed on the platform of Visual Studio 2012 with C# language as programming language.The aerial image mosaic system is simple to operate,which includes image mosaic,data loading,image zooming,distance area measurement,point,line and surface rendering,image enhancement,image stretching and other functions.