基于蚁群算法优化正射影像镶嵌线  

Optimizing ortho-image mosaic seamline based on Ant Colony algorithm

作　　者：马东岭[1] 丁宁[1] 崔健[1] 王京卫[1] 

机构地区：[1]山东建筑大学土木工程学院,济南250101

出　　处：《测绘科学》2013年第2期117-118,121,共3页Science of Surveying and Mapping

基　　金：住房与城乡建设部科技计划项目"基于数字摄影测量的城市三维景观建模及可视化研究"(2010-K9-24);住房和城乡建设部科技计划项目"基于3DGIS的古建筑物的三维重建与数字化保存研究"(2011-K9-25)

摘　　要：本文针对正射影像镶嵌线优化需求,本文提出了一种利用蚁群算法优化正射影像镶嵌线的方法。该方法将待镶嵌的相邻正射影像对根据地面坐标计算出它们在重叠区域的差值图像,通过外方位元素生成一条初始镶嵌线,并根据蚁群算法在重叠区域的差值图像上沿着初始镶嵌线在起点至终点选择一条最佳的路径避开房屋等高大地物,保存为最优镶嵌线。试验结果证明该算法能快速、有效选择正射影像的镶嵌线,实现大比例尺与城市地区的正射影像智能镶嵌。In connection with the optimized demand on ortho image mosaic seamline, a method of optimizing the ortho image mo- saic seamline using Ant Colony Algorithm was proposed in the paper. First, the images of different gray values between overlap areas of orthophotos were attained through difference operators. Then initial mosaic seamline was generated by exterior orientation elements. At last Ant Colony algorithm was used to select an optimal path that avoids the houses and other high terrain feature along the initial mosaic seamline from the starting point to the end point on difference images of overlapping area, and saved it to be the optimal mosaic seam- line. The experimental results showed that this algorithm could select seam line of orthophoto image by great speed and high accuracy, and realize DOM intelligence seamless mosaicking in surveying and mapping production.

关 键 词：数字正射影像 镶嵌线 蚁群算法 优化镶嵌线 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]