HYPERSPECTRAL REFLECTANCE SIGNATURES AND POINT CLOUDS FOR PRECISION AGRICULTURE BY LIGHT WEIGHT UAV IMAGING SYSTEM

Abstract

The objective of this investigation was to study the use of a new type of a low-weight unmanned aerial vehicle (UAV) imaging system in the precision agriculture. The system consists of a novel Fabry-Perot interferometer based hyperspectral camera and a high-resolution small-format consumer camera. The sensors provide stereoscopic imagery in a 2D frame-format and they both weigh less than 500 g. A processing chain was developed for the production of high density point clouds and hyperspectral reflectance image mosaics (reflectance signatures), which are used as inputs in the agricultural application. We demonstrate the use of this new technology in the biomass estimation process, which is based on support vector regression machine. It was concluded that the central factors influencing on the accuracy of the estimation process were the quality of the image data, the quality of the image processing and digital surface model generation, and the performance of the regressor. In the wider perspective, our investigation showed that very low-weight, low-cost, hyperspectral, stereoscopic and spectrodirectional 3D UAV-remote sensing is now possible. This cutting edge technology is powerful and cost efficient in time-critical, repetitive and locally operated remote sensing applications.