摘要
正射影像在影像边缘和覆盖有房屋、树木等地物的区域上表现出投影差,且投影差在不同的影像上不相同。当两幅正射影像镶嵌时,在重叠区域的差分影像上,这些区域表现为高亮度区,理想的镶嵌线应避开此类区域。采用改进的最小化最大算法在正射影像重叠区域的差分影像上自动搜索镶嵌线,其中采用贪心法取代Dijkstra算法进行局部路径选择,并且改进算法判定条件。试验表明,改进算法搜索的镶嵌线能够很好地避开投影差大的区域,且具有较好的自适应性。
Orthoimage has relief displacements in the edge area and area with buildings or trees.And displacements differ significantly on different orthoimages.An ideal seam-line should be away from high-displacement-area.An improvement of minimizing local maximum algorithm is proposed in searching seam-line on differential image of overlapped region.Improvements are mainly manifested in substituting Dijkstra algorithm for Greedy algorithm in local path selection and modifying decision criteria.The result shows that improved method is self-adaptive,and is able to select a seam-line away from high-displacement-area,which fulfills the request of seamless mosaic.
出处
《测绘学报》
EI
CSCD
北大核心
2012年第2期199-204,共6页
Acta Geodaetica et Cartographica Sinica
基金
国家创新研究群体科学基金(41021061)
中央高校基本科研业务费专项资金(201121302020009)
关键词
正射影像镶嵌
贪心搜索法
最小化最大算法
orthoimage mosaicking
greedy select algorithm
minimize local maximum algorithm
