Font Size: a A A

Study On The Key Technologies Of True Orthophoto Generation Of Unmanned Aerial Vehicle Image

Posted on:2018-03-11Degree:MasterType:Thesis
Country:ChinaCandidate:Y F XieFull Text:PDF
GTID:2310330518459450Subject:Surveying and Mapping project
Abstract/Summary:PDF Full Text Request
Aerial Image has a rich texture,realistic color,a strong intuitive.It is easy to update with a good current potential.It can be used for digital mapping,and used as a base map to update the database.The ordinary orthophoto does not take the surface of the building into consideration,so the image of the building will appear to tilt and occluded.In the process of image mosaic,it is difficult for ordinary orthophoto to splice the image.With the continuous improvement of imaging techniques,the image resolution we acquire is also getting higher and higher.In the tall-building-intensive area,the problem of tilt and occlusion of building is particularly obvious.When the buildings are tilted and occluded,the orthophoto is difficult to apply to other map sets,which can make the orthophoto lose the geographical reference value and can not be used as the base map of the database.True orthophoto can solve the problems mentioned above.The difference between true orthophoto and the ordinary orthophoto is that true orthophoto takes into account the projection error,which can correct the roof of the building to its corresponding ground position,and solve the problem of occlusion of buildings and splicing difficult problem.True orthophoto has both the visual features of the image and the characteristics of the topographic map,so it is significant to study the true orthophoto production.True orthophoto uses digital surface model in digital differential rectification.The accuracy of the orthophoto depends largely on the accuracy of the digital surface model.The digital surface model not only describes the surface morphology but also the elevation information of the ground building,which can be regarded as the combination of the digital elevation model and the digital building model.Therefore,the precise DEM and DBM are the prerequisites for constructing DSM.Taking the UAV images of Chengdu University of Technology as the data source,and taking the tall and dense building area of the school as an example,This paper studies the key technologies of true orthophoto production.The main research contents are as follows:(1)The point cloud data of experimentation area were extracted by image matching technology.In order to obtain the ground point and obtain accurate DEM,this paper explored the DEM filter.For the problem that automatic point cloud filtering can not obtain accurate ground points,artificial fine point cloud editing is adopted.Taking the building of Dongyuan as an example,this paper summarized a set of DEM fine editing program for tall-building-intensive area.(2)Based on the original digital surface model and the edited DEM producing digital orthophoto,the existing deformation and tilt of the building are analyzed.This paper selected the images of the building projection error at the minimum position based on the orthorectification of edited DEM for mosaicing and optimized the seamlines,which solved the problem of tilt and occlusion of buildings.In order to solve the problem of building tilt completely,this paper constructed a digital model describing the surface of a building—Digital Building Model(DBM),and reconstructed the digital surface model in the form of "DEM + DBM",which can describe buildings and surface morphology more accurately than the original digital surface model.(3)This paper studied the "repetitive image" problem based on DSM orthorectification,and realized parallel processing of the image orthorectification and occlusion area detection.This paper summarized a set of seamlines optimization scheme based on texture compensation for the complexity of the true orthophoto seamlines,and finally got a seamless splicing of the true orthophoto,which solved the problems very well that the building are tilted and occlude each other in the ordinary orthophoto images.
Keywords/Search Tags:Image processing, True orthophoto, Occlusion detection, DSM, DBM, Texture compensation
PDF Full Text Request
Related items