基于车载二维激光扫描的树冠体积在线测量

采用车载二维激光扫描仪获取树木单侧点云数据,坐标变换后通过设置感兴趣区域检测树木,利用垂直分布特性识别树干,得到树冠中心距离。考虑树冠连续/不连续2种情况进行树木分割,将树冠外缘距离与对应树干距离相减算出树冠厚度。将树冠体积离散化为长方体,利用树冠厚度、相邻测量点垂直方向距离、车辆速度、扫描周期等参数进行计算。采用FIFO缓冲区保存在线数据,新采集的一帧数据写入缓冲区末尾,同时从缓冲区开头读出处理好的数据帧输出,实现树冠体积的在线测量。实验结果证明,树冠连续/不连续场景下,方法均能准确检测分割树冠、识别树干,实现树冠体积的在线测量。

Online Measurement of Tree Canopy Volume Using Vehicle-borne 2-D Laser Scanning

An online measurement method of tree canopy volume using vehicle-borne 2-D laser scanning was proposed. Firstly, the point cloud data of trees was obtained by using a vehicle-borne 2-D laser scanner. After coordinate transformation, trees within the region of interest were detected. Then trunks were recognized by using their vertical distribution characteristic and trunk distances were obtained. In the case of canopy continuity, tree segmentation was performed by finding minimum data frame between trunks. Crown thicknesses were computed by subtracting the crown outer edge distances from the corresponding trunk distance. Finally, tree canopy was discretized into small boxes and its volume was computed by using crown thickness, distance between adjacent measuring points in the vertical direction, vehicle speed and scanning circle. To meet the demand for online measurement, data was saved in FIFO buffer which included all information required for the calculation of tree canopy volume. New data frame was written to the end of the buffer and the processed data frame was read from the beginning of the buffer. The experimental results showed that in the continuous and discontinuous canopy scenarios, trees and trunks were detected effectively and tree canopy volume was measured online accurately. In the experiment, the average processing time per frame was 2.0 ms and the maximum processing time was 15.6 ms, which was less than laser scanning period (25ms) and met the needs of real-time processing.