STPAS: Spatial-Temporal Filtering-Based Perception and Analysis System for Precision Aerial Spraying

Abstract

This study proposes a perception and analysis method for precise aerial spraying based on three-dimensional (3D) deep learning. Point cloud data for water droplets are acquired using 3D LiDAR, and the PointNet++ deep learning model is trained to classify and segment the spray pattern. Spatial-temporal data are processed for the segmented point cloud data. The spray from each nozzle is clustered through spatial data processing, and clustering is based on this information. This approach allows each nozzle to be distinguished and mapped. Processing temporal data compensates for unsensed or noisy data points and predicts the water droplet trajectories, enhancing the spray data. This method more accurately measures the shape of water droplets. Experiments altering the flight conditions of unmanned aerial vehicles (UAVs) were conducted to assess the proposed framework, demonstrating that processing is feasible in the onboard system of the UAV. The proposed method has potential application in control systems for precise spraying in the future.