激光雷达点云数据的飞机残骸拼接重建  

LIDAR point cloud data-based mosaic reconstruction of aircraft wreckage

作　　者：孙华波 罗博仁 俞力玲 刘慧丽[2] 钟民主 晏磊[2] 

机构地区：[1]中国民航科学技术研究院航空安全技术实验室,工程师北京100028 [2]北京大学空间信息集成与3S工程应用北京市重点实验室,北京100871

出　　处：《中国安全科学学报》2014年第3期109-113,共5页China Safety Science Journal

基　　金：国家自然科学基金资助(U1333124,61039002);民航安全能力建设项目(132146903054)

摘　　要：为查明航空事故的原因,需拼接重建飞机残骸。提出一种用激光雷达(LIDAR)采集飞机残骸点云数据的方案,并构建残骸拼接模型。在实际操作中,对机身残骸采用改进的迭代就近点(ICP)拼接算法进行精确处理;对零散残骸采用选取同名点拼接算法进行处理。试验结果显示,使用该模型和改进迭代就近点拼接算法处理采集的多站数据,能够实现飞机残骸的拼接重建,同时将拼接精度提高3倍。研究结果表明,与物理拼接相比,采用LIDAR点云的飞机残骸拼接重建,可以缩短事故调查的周期。For identifying causes of aviation accidents,reconstruction and splice of wreckage are a must. This paper contributes to this research by working out a wreckage splice method based on LIDAR point cloud data and building a model for wreckage reconstruction. According to the wreckage of fuselage and components,describe their reconstruction methods were described respectively. An improved ICP algorithm was used to process main body of aircraft. The same name point stitching algorithm was used to process components wreckage. Experiment results show that both the model and improved algorithm can be well used to complete the mosaic reconstruction of aircraft wreckage,and that accuracy is improved by 3 times. The research results show that compared with physical splicing,the LIDAR point cloud databased method will be helpful in shortening accident investigation periods.

关 键 词：航空运输 残骸拼接 改进迭代就近点(ICP)算法 激光雷达(LIDAR)点云 事故调查 

分 类 号：X928.02[环境科学与工程—安全科学]