论文部分内容阅读
植物表型研究在把握植物结构和了解植物对环境变化的反应上起着至关重要的作用。因此,通过引入多种遥感技术例如LIDAR,研究者一直针对不同场景开发不同的植物表型技术。目前,该领域的重要趋势是开发可供于广泛使用的低成本三维系统,而通常使用的高成本的LIDAR传感器无法满足这一需求。为了解决这个问题,试图开发一个基于低成本二维激光扫描仪的用于室内植物三维结构表型研究的移动地面近端感应系统。特别地,将两个RPLIDAR激光扫描仪(目前市场上最低成本的二维LIDAR传感器)安装在两个远端的放置于移动平台的固定架上。如此一来,它们的扫描剖面被设置成一个斜交的方式。然后,经过精确的数据三维位置坐标设定和校准后,平台的移动可以使得两个系列的二维扫描剖面转变为研究区一排植物的全部三维表征。基于三维点云结果,人们可以获得详细的植物结构特征。实验表明,提出的方案得到基本验证,对于特定的植物结构变量例如叶面积(R2=0.92)可以精确获取。总之,该工作可以推动将基于LIDAR的植物表型技术发展到真正意义上的低成本阶段,而且这表明LIDAR在植物表型研究上的实际应用会越来越多出现在植物栽培、精准农业等领域。
Plant phenotypic research plays a crucial role in understanding plant architecture and understanding how plants respond to environmental changes. Therefore, by introducing various remote sensing technologies such as LIDAR, researchers have been developing different plant phenotyping techniques for different scenarios. At present, the important trend in this field is to develop a low-cost three-dimensional system that can be widely used, and the commonly used high-cost LIDAR sensor can not meet this requirement. In order to solve this problem, an attempt was made to develop a mobile ground near-end sensing system based on the low-cost two-dimensional laser scanner for studying the three-dimensional structure of indoor plants. In particular, two RPLIDAR laser scanners (the lowest cost two-dimensional LIDAR sensor on the market today) are mounted on two remote mounting brackets placed on a mobile platform. As a result, their scanning profiles are set in a skewed manner. Then, after accurate three-dimensional positional coordinate set-up and calibration of the data, the movement of the platform allows the two series of two-dimensional scan profiles to be transformed into the full three-dimensional characterization of a row of plants in the study area. Based on the 3D point cloud results, one can obtain detailed plant structural features. Experiments show that the proposed scheme is basically validated and can be accurately obtained for specific plant structural variables such as leaf area (R2 = 0.92). All in all, this work can promote the LIDAR-based plant phenotyping technology to a real low-cost stage, and this shows that the actual application of LIDAR in plant phenotyping will be more and more in plant cultivation, precision agriculture, etc. field.