WBD-YOLO-AM: YOLOv8 with attention module and IoT-based wild boar detection and deterrence system for safeguarding small millets
Year: 2025
Type: article
Abstract: Small millets are traditional and climate-resilient food crops that can grow in adverse weather conditions. Animal intrusion, specifically from wild boars, is a main threat to the production of small millets, which reduces interest in the cultivation of small millets and affects farmers’ incomes. To address the problem of wild boar attack on agriculture land, an improved YOLOv8 is proposed, with an attention module based on WBD-YOLO-AM and that works from the collection of data, its preprocessing, attention mechanism to improve the feature maps, training then building the model, after that the models hyperparameter are tuned to boost the performance of the model and it outperforms the state-of-the-art approaches and achieves the impressive rate of 97.1% precision, 96.2% recall and 98.2% mAP. After attaining ideal results, a Raspberry Pi (Raspberry Pi Foundation, Cambridge, UK) equipped with a camera records wild boar images in real time. The Raspberry Pi is further attached to a speaker that emits ultrasonic sound to deter off wild boar. After training, the enhanced YOLOv8 model is installed on a Raspberry Pi for software integration. It can detect wild boars with accuracy and plays a deterrent sound, a Raspberry Pi and a GSM module is used to send an alarm. The importance of this research lies in its potential to protect agricultural land from wild boar damage without causing any human-wild life conflict.
Authors: ADITYA JOSHI, NEHA PANDEY, MANOJ DIWAKAR, PRABHISHEK SINGH, ACHYUT SHANKAR, FAYEZ ALQAHTANI
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDGs: Zero hunger, Life on land
Open Access status: hybrid
Funder: King Saud University
OpenAlex ID: W4413771960