Recently, a joint research team from the National Engineering Research Center for Agricultural Informatization, the Information Technology Research Center of Beijing Academy of Agriculture and Forestry Sciences, and the Xi'an Institute of Optics and Precision Mechanics of the Chinese Academy of Sciences has successfully developed China's first domestic new video hyperspectral and point cloud "image-spectrum integration" sensor. This technological breakthrough fills the gap in domestic related fields and provides powerful technical support for crop breeding and precision agriculture, especially for agricultural drone applications. The sensor adopts a modular design, integrating hyperspectral collection and 3D point cloud measurement, which is lightweight and agile, suitable for being carried by small and light agricultural drones and ground mobile platforms.
It achieves three major technological breakthroughs: first, it uses domestic pixel-level spectral filtering snapshot array spectroscopy technology to realize video-level high-speed synchronous imaging, overcoming the problems of large geometric distortion and difficult correction of traditional linear push-broom hyperspectral imagers; second, it integrates an advanced domestic high-precision surveying and mapping-level lidar, which can generate 200-300 laser profiles per second, with a point density 4-6 times that of traditional lidar and an absolute spatial position accuracy better than 3 centimeters; third, it realizes high-precision image-spectrum fusion of 3D point clouds and hyperspectral images on mobile platforms such as drones, with a geometric measurement accuracy better than 1 pixel. This sensor enables agricultural drones to quickly obtain millimeter-resolution imaging hyperspectral data and synchronous fine 3D morphological data of crops, realizing real-time synchronous perception of crops in five dimensions: space, three-dimensional, spectral and time, which is of great significance for promoting the intelligent upgrading of agricultural drones.
