基于图割法的无人机正射影像拼接

UAV Orthoimage Mosaic Based on Graph Cut

无人机正射影像具备精确的几何定位和丰富的地物信息,在地理信息系统等领域发挥着重要作用,将像幅较小的单幅影像进行拼接是数据发布前必不可少的操作。目前拼接算法因没有考虑图像局部信息,不能较好的保证拼接后图像地物的完整性。针对上述问题,本文采用图论中的图模型优化问题,从亮度值、梯度值以及纹理复杂度三方面对有向边进行定权,采用图割法中最大流-最小割算法进行全局求解。实验表明该算法真实有效,且生成的拼接线可以较好的避免穿过房屋建筑,进而较好的保证了地物的完整性。

Digital orthoimages obtained by unmanned aerial vehicle play an important role in geographic information systems and other fields,because of the accurate geometric positioning and rich ground information.It is an indispensable operation before data release to mosaic a small image Without considering the local information of image,current algorithms the integrity of the ground objects after mosaicing cannot be better guaranteed.In view of the above problems,this paper used the graph model optimization problem in graph theory.By determining the weights of the directed edges from three aspects of brightness value,gradient value and texture complexity,then used the max flow-min cut algorithm in the graph cut to solve the problem globally.Experiments shown that the algorithm was real and effective,and the generated seamline can better avoid passing through the building,and thus better ensured the integrity of the ground objects.