Research On UAV Images Ortho-rectification And Mosaic Technology And Its Application

In recent years,as people pay more and more attention on the research and development of rapid mapping and high-precision mapping,compared with the unavoidable factors such as the strong impact which is vulnerable to the external environment,time-consuming operation,poor flexibility and etc.showed by the traditional space remote sensing platform,UAV with its unique advantages such as high mobility,flexibility,high efficiency,low consumptionand etc.,not only successfully becomes a reasonable substitute of traditional remote sensing platforms and conventional aerial photogrammetry techniques,but also is brought into closely focus and researched by related industry.Especially UAV has obvious advantages in the small range of high-precision positioning data and high score image fast access,disaster emergency response and dangerous areas of air monitoring and so on,and in the face of geographical census,national emergency relief and digital city construction and other aspects of the development of UAV images for fast processing algorithms,has a very critical advantage.In the low-altitude UAV remote sensing image processing system,the orthorectification and splicing of the image is the key to the remote sensing image processing of the unmanned aerial vehicle.For the realization of digital orthophoto imaging automation production,it is important to choose a suitable and efficient method of the UAV aerial image orthophoto image correction and splicing.The emphasis of this paper is the orthogonality correction and stitching technology of UAV images,and the article focuses on the following aspects:(1)This paper discusses in detail the basic principle of image ortho-rectification.It studies the principle and processing flow of several ortho-rectification methods,and analyzes the advantages and disadvantages of these methods and compatibility in different conditions;(2)This paper discusses the splicing method of UAV orthophoto image.It describes the principle of image mosaic algorithm,the operating process,advantages and disadvantages of Dijkstra algorithm,splicing line of dynamic programming method,Twin-snake algorithm mosaic line detection and ant colony algorithm mosaic lines;(3)Combined with Changchun Jingyue and Dehui City Tongtai Township Weizi Village project examples,This paper validates and analyzes the experiment of orthogonally correcting and splicing of UAV images.Jingyue experiment area is mainly composed of hilly areas and residential areas,UAV carrying Sony a7 r camera intakes of two voyages,59 photography images,range of area: 43°42?01.6488??N-43°42?08.23824??N,125°19?14.8314??E-125°20?17.7738??E,and the direction of the flight is roughly eastward;Terrain of Dehui City Tongtai Township Weizi Village is flat,mainly is the plain terrain,also contains some residential areas,UAV carrying Sony ILCE-QX1 camera intakes of 20 voyages,884 photography images,range of area: 44 ° 23 ? 09.276 ??N-44 ° 25 ? 58.8 ??N,125°20?46.068??E-125°25?20.928??E,and the direction of the flight is roughly north-south.This paper analyses aerial triangulation of UAV images combined with POS data in two experimental area,corrects the single image one by one,and finally stitch all orthophoto images.From the experimental area of Dehui City Tongtai Township Weizi Village we select 100 checkpoints on the processing of ortho-rectification images for precision testing,and get the result of the mean square error of a point is 12.2cm,the result meets the accuracy requirements;This paper selects local area superimposed with DLG and analyse the accuracy of orthophoto images,it is found thatthe residential area and the plain area gets the highest quality image,hilly areas followed.