林木参数监测多传感器的时空配准

Spatial-temporal Registration of Multi-sensor for Forest Parameter Monitoring

本文基于多传感器集成以及时空配准技术,实现林木的立体参数采集。采用自主研发的林木参数多传感器集成系统,在东北林业大学林场样地进行林木的数据采集,利用传感器间坐标系转换关系与双目视觉算法实现点云数据和图像数据的时空配准。对配准后数据提取标靶坐标与全站仪得到真实坐标进行比较,X、Y、Z三个方向的RMSE分别为0.074、0.117、0.153m,RMSE平均值为0.115m。结果表明该集成系统采集数据有效,配准效果良好。

Based on multi-sensor integration and spatial-temporal registration technology,this paper realizes the stereo parameter acquisition of forest trees.The self-developed forest parameter multi-sensor integrated system is used to collect the data of forest trees in the forest farm sample of Northeast Forestry University,and the spatial-temporal registration of point cloud data and image data is realized by the coordinate transformation relationship between sensors and the binocular vision algorithm.After the registration,the data extraction target coordinates are compared with the total station to obtain the real coordinates.The RMSE in the three directions of X,Y and Z are 0.074,0.117,0.153 m,respectively,and the average RMSE is 0.115 m.The results show that the data collected by the integrated system is valid and the registration effect is good.