Ramírez, Germán (Centro de Investigaciones en Óptica), Montes de Oca Rebolledo, Andres (Centro de Investigaciones en Óptica), Flores, Gerardo (Centro de Investigaciones en Óptica)

3D Maps of Vegetation Indices Generated Onboard a Precision Agriculture UAV

Scheduled for presentation during the Regular Session "UAS Applications III" (ThA5), Thursday, June 8, 2023, 09:00−09:20, Room 466

2023 International Conference on Unmanned Aircraft Systems (ICUAS), June 6-9, 2023, Lazarski University, Warsaw, Poland

This information is tentative and subject to change. Compiled on April 26, 2024

Abstract

Vegetation indices are qualitative indicators that permit the identification of damaged or healthy zones within a vegetation region. Traditionally, the vegetation indices are represented as 2-dimensional maps since they are computed from aerial imagery. However, 3D mapping is more convenient when a detailed study of a plant is required. This work proposes an Unmanned Aerial System that can provide indices for low-height vegetation. Such an output is presented in a dense point cloud format. The proposed system features an NVIDIA Jetson AGX Xavier as the onboard computed and a ZED Mini stereo as the imaging sensor. Since the ZED Mini captures RGB imagery in the visible spectrum, we focus on computing two visible-based vegetation indices: the Green Normalized Difference Vegetation Index and the Visible-band Difference Vegetation Index. We present the results of two experimental flights to prove the system’s functionality. In such flights, we perform an online reconstruction using the RTABMap algorithm. After obtaining the dense point cloud of the vegetation regions, our system processes the data to output 3D maps of the vegetation indices above. Additionally, it is presented as a simple quantitative analysis that can be used for vegetation segmentation applications.