Unmanned aerial vehicles (UAVs) find astounding applications in some of the key areas of search and rescue operations, precision agriculture (PA), remote sensing, etc. The unique ability of a UAV to hover right above the target location and collect useful information makes it a suitable candidate for many such applications. UAV has been utilized in many key applications such as navigation and control, and visual-based detection, but autonomous decision-making and carrying out mid-air actions, such as PA for weed control, is an open research area. In the proposed study, a novel framework has been presented for detecting multiple targets (crop, weed, mud, etc.), recognition and navigation, and subsequent movement of UAVs towards the proposed targets. The successful implementation of the model in UAV would enable it for weed control in PA, among others. The presented framework utilizes a robust computation, convolutional neural network (CNN) model for target detection and recognition, while proximate probing is performed using a PID-based algorithm, ensuring UAV hover right above the target. The proposed framework has been developed in five successive steps by adopting Lawson’s sense, process, compare, decide, and act model. The framework is successfully validated on simulation flights conducted in Gazebo and PX4 and real field experiments. The experiments’ results showed the versatility of the developed system to many complex missions where targets are added or removed.
An Intelligent Vision-Guided Framework of the Unmanned Aerial System for Precision Agriculture
Algorithms for Intelligent Systems
Computer Vision and Machine Learning in Agriculture, Volume 3 ; Chapter : 12 ; 159-176
2023-08-01
18 pages
Article/Chapter (Book)
Electronic Resource
English
Intelligent pesticide spreading unmanned aerial vehicle for intelligent agriculture
European Patent Office | 2025
|Special intelligent unmanned aerial vehicle for agriculture and forestry
European Patent Office | 2015
|Pesticide spraying unmanned aerial vehicle based on intelligent agriculture
European Patent Office | 2022
|