ENHANCING THE PERFORAMANCE OF AN EXISTING IOT BASED SMART ANIMAL DETERRENT SYSTEM FOR CROP PROTECTION USING CORAL USB ACCELERATOR

Crop protection remains a serious challenge for modern agriculture, particularly in mitigating damage caused by wild animals. This project presents a smart animal deterrent system with a Coral USB Accelerator to enhance real-time detection and response capabilities. A coral USB device was integrated into an existing work carried out by a group of students in the previous session. The existing system can accurately and efficiently identify various animal species in real-time, even under challenging environmental conditions. When potential threats are detected, the system activates deterrence mechanisms to prevent crop damage, ensuring a timely and effective response. It utilises conventional sensor networks and basic image processing. The upgraded system employs advanced computer vision techniques powered by deep learning models optimised for the Coral USB Accelerator. This hardware improvement facilitates fast processing of high-resolution images, reducing inference latency and lowering power consumption. The system demonstrated that the addition of the Coral USB Accelerator led to a notable improvement in detection accuracy and speed compared to previous iterations. The enhanced system exhibits robust performance in dynamic agricultural settings, offering a scalable and cost-effective solution for safeguarding crops. This work contributes to knowledge, the benefit, of combining edge computing with machine learning to address critical challenges in crop protection.