Edge-IoT-UAV Adaptation Toward Precision Agriculture Using 3D-LiDAR Point Clouds
Arun Kumar Sangaiah, Jayakrishnan Anandakrishnan, Venkatesan Meenakshisundaram, Mohd Amiruddin Abd Rahman, Padmapriya Arumugam, Mrinali Das
Abstract
Precision agriculture significantly boosts socio-economic growth and national productivity through monitoring accurate periodic biomass and biophysical traits. Numerous Internet of Things (IoT) and Unmanned Aerial Vehicles (UAVs) connected sensors can facilitate the automated collection of these traits, even in adverse conditions. This article introduces SmartAgri-Net (SA-Net), a decision support system that utilizes real-time multi-sensor and multi-temporal data from Edge-IoT-UAV sensors. The SA-Net-Biomass Estimation Framework (SA-BEF) consisting of Occlusion Reconstruction Module (ORM), 3D–2D Transfer Block (3-2DTB), Attention-based Biomass Estimation Block (ABE) approximates biomass from Light Detection and Ranging (LiDAR) 3D-Point clouds. The SA-Net-TCN-Prediction Framework (SA-TPF) implements Temporal Convolution Neural Network (TCN) for predictive analytics over the derived biomass and aggregated biophysical data from IoT sensors to perform decision-making. Finally, we propose engineering and deploying SA-Net recom-mendation support for smartphone applications.