基于影像滤除地面激光雷达数据中绿色植被点云的方法研究  

Filtering Method of Green Vegetation in Terrestrial Laser Scanning Point Cloud Data Based on Image

作　　者：汪沛[1] 宋海峰[2] 赵迎辉[3] 赵睿[1] 

机构地区：[1]北京林业大学理学院,北京100083 [2]中国电子科学研究院,北京100041 [3]中国科学院遥感与数字地球研究所,北京100094

出　　处：《安徽农业科学》2014年第16期5309-5312,共4页Journal of Anhui Agricultural Sciences

基　　金：中央高校基本科研业务费专项(YX2013-34)

摘　　要：基于地面激光雷达扫描时配置的高分辨率相机所获取的影像数据,变换影像数据的颜色空间之后分别在色相通道和饱和度通道中根据阈值检测出绿色植被的分布区域,通过融合两个通道的阈值检测结果提高检测准确度,通过对检测区域进行腐蚀和膨胀运算消除检测区域中的数据噪点,最终通过二维检测区域和三维点云数据的配准实现滤除绿色植被所对应的点云数据。利用地面激光雷达实际测量的点云数据和影像数据开展试验,获得了较好的试验结果,表明方法有效可行。Based on the image data obtained by a high-resolution camera mounted on the terrestrial LiDAR,the color space was changed,then the hue and saturation information were used to identify the distributed area of green vegetation.The extracted green vegetation areas were processed by the erosion operation and expansion operation to eliminate the noise.Then the images marked with the extracted green vegetation areas were registered with the point cloud data,and be filtered out according to the marked areas.Finally,the practical point cloud data and images acquired by the terrestrial LiDAR,were tested with the proposed method.The good results demonstrated that the method was verified to be feasible and effective.

关 键 词：地面激光雷达 点云 绿色植被 滤除 

分 类 号：S126[农业科学—农业基础科学] TN958.98[电子电信—信号与信息处理]