Map Based On Aerial Video Reconstruction

Environment building based on image modeling and rendering is the hot point of theresearch on computer vision reality technology. In this paper, I mainly do some research onhow to mosaic images sequences acquired in the air, whose position and stance are easilyeffected by wind and not stable, we should find a solution to obtain the altitude of the camerawith the computer vision technology.The approach to image registration is studied first, according to the chararctes of theimages acquired, a quick and stable algorithm is developped. Camera calibration is the essentialstep to obtain the information ofthree-dimensional structure from two-dimensional images inthe field of computer vision. We use knowledge and technique around the theory of cameracalibration. to obtain standard image, which means camera vertically placed when imagesbeing taken. Images acquired by the same camera in different orientation on the same planarare related by a homography matrix. So all images can be placed into the standard imagecoordinates and piece together as a bigger image. The big image is the 2-Dimension map wetry to obtain.By designing the experiment scientifical based on the image sequences, we analyze thevirtue and shortage of several approachs for each steps. In the last chapter, a whole experiment isdone to show the whole procedure for reconstruction.The paper involves knowledge about image registration, camera calibration image mosaic, and soon. It gave a tentative research on the intercourse between Photograrm and computer vision.Experiments results shows that the methods have some feasibility in real application.