利用Kinect进行地面植被分层覆盖度估算研究

Estimating fractional vegetation cover at different heights via Kinect

植被覆盖度(fractional vegetation cover,FVC)是一个重要的植被生物物理学参数,是一定空间尺度下所有植被覆盖比例的综合反映.植被垂直方向的异质性在很多应用中也需要考虑.目前遥感和地面测量对获取植被垂直结构都有待完善,新仪器、新方法仍在不断发展之中.本文采用一种商业化的3D体感摄影机——Kinect,其传感器具有彩色和深度镜头,相当于数码相机和激光雷达的结合.实验发现,Kinect能够用于获取空间相匹配的数码相片和深度图像,并得到植被的垂直剖面信息,进而计算垂直分层的FVC.受限于Kinect目前的有效探测范围,这种设备不能大规模地应用于野外测量中,但是它可为发展结合激光雷达和传统光学传感器获取植被参数的新方法提供测试数据.

Proportion of vegetation cover per area, the fractional vegetation cover （FVC）, is an important parameter in ecology and forestry studies. Trees, shrubs and undergrowth or short vegetation often contribute to estimated FVC. It is often difficult to distinguish fractions of each type. However, some knowledge of vertical distribution of vegetation is frequently useful. New equipment and new methods could often improve works in remote sensing and field measurements. A new equipment, Kinect, was used here to acquire vertical FVC information. Kinect is a motion-sensing device launched in 2010, composed of a red-green-blue （RGB） camera, an infrared laser emitter and a receiver. Kinect could be used to obtain RGB images and depth images simultaneously, resembling a combination of LiDAR （laser intensity direction and ranging） and digital camera. Depth and spectral information from Kinect could be used to describe vertical vegetation distribution and to estimate FVC on height parameter. Color information from RGB images and depth images could be used respectively for vegetation classification and FVC layering. Although there are still some limitations in the detecting range of Kinect, our experiments have shown potential of combining LiDAR and traditional optical sensors in field vegetation measurements.