欧建良, 鲍峰, 王卫安, 陆娇. 地面移动测量近景影像的建筑物道路特征线段自动分类研究[J]. 武汉大学学报 ( 信息科学版), 2011, 36(1): 60-65.
引用本文: 欧建良, 鲍峰, 王卫安, 陆娇. 地面移动测量近景影像的建筑物道路特征线段自动分类研究[J]. 武汉大学学报 ( 信息科学版), 2011, 36(1): 60-65.
OU Jianliang, BAO Feng, WANG Weian, LU Jiao. Building and Road Characteristic Line-segment Automatic Classification Research of Ground Mobile Mapping Close-range Imagery[J]. Geomatics and Information Science of Wuhan University, 2011, 36(1): 60-65.
Citation: OU Jianliang, BAO Feng, WANG Weian, LU Jiao. Building and Road Characteristic Line-segment Automatic Classification Research of Ground Mobile Mapping Close-range Imagery[J]. Geomatics and Information Science of Wuhan University, 2011, 36(1): 60-65.

地面移动测量近景影像的建筑物道路特征线段自动分类研究

Building and Road Characteristic Line-segment Automatic Classification Research of Ground Mobile Mapping Close-range Imagery

  • 摘要: 根据影像灭点理论分析建筑物道路特征线与影像平面线段间数学关系,对影像所提取线段进行分类。实验通过立体影像中模拟数据和基于影像边缘提取线段的处理实现了地物特征线段的快速自动分类,无关线段过滤效果显著,是移动测量地物自动量测与影像理解的重要基础。

     

    Abstract: Ground mobile mapping system(MMS) is a brand-new integration system in survey and mapping field.It collects digital measurable real imagery for geo-objects(e.g.building and road) with multi-and ground-views easily.MMS works in a lower cost and higher speed way compared to other survey methods,and its imagery data contains rich information for observed objects.It becomes a fundamental issue for the observed object characteristic line-segments to be classified from MMS imagery accurately and rapidly.The paper presented a new method that including mathematic model between building and road 3D characteristic lines and MMS imagery with vanishing point theory,and imagery line-segment classification and filtering by vanishing points.Experiment result shows that object characteristic line-segments were obtained completely while other noise line-segments excluded accurately and the total number of relevant imagery line-segments is reduced markedly.Thus,an important basic work of automatic object measuring and imagery understanding is finished in the paper.

     

/

返回文章
返回