DSM辅助下城区大比例尺正射影像镶嵌线智能检测解析.PDFVIP

40 1 Vol.40, No. 1 20 11 2 Acta Geodaetica et Ca rtographica Sinica Feb. , 20 11 : 1001-1595 (2011) 01-0084-06 DSM 左志权, 张祖勋, 张剑清, 曹 辉 , 4 30079 Seamlines Intelligent Detection in a rge-scale Urban Orthoimage Mosaicking ZUO Zhiquan, Z HANG Zuxun, Z HANG Jia nqing, CAO Hui School of Remote Sensing a nd Information Engineering, Wuhan University, Wuha n 430079, China Abstract : Ground regional a nd non-ground area a re a ccurately distinguished by processing high-precision DSM. Most of tra ditional seamlines detection algorithms are based on iterative calculations, and have highcomplexity. It is proposed a new algorithm na med greedy sna ke algorithm. T he new a lgorithm is only associated with three pa rameters: sea rching step, direction rotation interval and region width. Experiments for urban areas for orthopho- to seamlines detection show that the sea mlines ca n avoid buildings well, a nd the greedy snake algorithm can be applied in the optima l path detection effectively. Key words : orthophoto mosa ic; image segmentation; topology reasoning; greedy snake; seamline detection :以高精度数字表面模型为检测基础, 通过智能识别 理手段对地面区域与非地面区域进行良好区分, 发展出一 种贪婪蛇型算法该算法仅与搜索步距方向旋转间隔以及重叠区域宽度等三 个因素相关, 具有速度快效率高等优点 针对典型的城市区域进行镶嵌线检测试验, 表明DSM辅助下的城区正射影像镶嵌线能很好地避开房屋等明显地表实 体,并且贪婪蛇型算法能有效地进行最优路径检测 :正射影像镶嵌; 影像分割; 拓扑推理; 贪婪蛇型算法; 拼接线检测 : P237 : A : 国家863 计划(2009AA12Z126 ,2009AA12 140 3-2 ( 1)) 1 引 言 , : , : ¹ , Pix el Facto ry , ; º Inpho SSK, ( ) 90% , , , , 1] RG B H IS , , , ( t w in snakes)