IoT-based real-time object detection system for crop protection and agriculture field security
Priya Singh,Rajalakshmi Krishnamurthi
DOI: https://doi.org/10.1007/s11554-024-01488-8
IF: 2.293
2024-06-05
Journal of Real-Time Image Processing
Abstract:In farming, clashes between humans and animals create significant challenges, risking crop yields, human well-being, and resource depletion. Farmers use traditional methods like electric fences to protect their fields but these can harm essential animals that maintain a balanced ecosystem. To address these fundamental challenges, our research presents a fresh solution harnessing the power of the Internet of Things (IoT) and deep learning. In this paper, we developed a monitoring system that takes advantage of ESP32-CAM and Raspberry Pi in collaboration with optimised YOLOv8 model. Our objective is to detect and classify objects such as animals or humans that roam around the field, providing real-time notification to the farmers by incorporating firebase cloud messaging (FCM). Initially, we have employed ultrasonic sensors that will detect any intruder movement, triggering the camera to capture an image. Further, the captured image is transmitted to a server equipped with an object detection model. Afterwards, the processed image is forwarded to FCM, responsible for managing the image and sending notifications to the farmer through an Android application. Our optimised YOLOv8 model attains an exceptional precision of 97%, recall of 96%, and accuracy of 96%. Once we achieved this optimal outcome, we integrated the model with our IoT infrastructure. This study emphasizes the effectiveness of low-power IoT devices, LoRa devices, and object detection techniques in delivering strong security solutions to the agriculture industry. These technologies hold the potential to significantly decrease crop damage while enhancing safety within the agricultural field and contribute towards wildlife conservation.
computer science, artificial intelligence,engineering, electrical & electronic,imaging science & photographic technology