CluckTrack: Manok Bisaya Monitor using Computer Vision Yolo V8 and IOT GSM Alert for Sustainable Farming

CluckTrack is an advanced system designed for real-time surveillance and management of chickens within a coop environment. It utilizes the YOLOv8s algorithm for object detection, leveraging an ESP32-CAM to capture images and videos, which are transmitted via WiFi to a Firebase database for processing and storage. The ESP32-CAM is powered directly by a power bank. An Arduino Uno, connected directly to a computer, manages the SIM800C GSM module, facilitating cellular connectivity for remote monitoring. In the event of a missing or inactive chicken, Firebase triggers a notification to the GSM module, which sends an SMS alert to the user. The system was evaluated using a custom dataset of 118 annotated images of Manok Bisaya chickens, demonstrating impressive performance metrics, including a precision of 92.8%, recall of 93.9%, mean Average Precision (mAP) of 95.3%, overall accuracy of 82%, and an F1 score of 93.3%. The YOLOv8s algorithm showed satisfactory performance with minimal missed detections, averaging an inference time of 0.9739 seconds per image. The Flutter-based app offers functionalities such as viewing the image gallery, updating contact information, checking the number of chickens in the coop, and receiving notifications. Recommendations for future improvements include expanding the dataset, upgrading the camera and microcontroller, integrating single-board computers for on-device processing, and adding WiFi configuration capabilities in the app. CluckTrack provides a robust solution for farmers, enabling efficient management of their flocks through advanced image recognition, real-time notifications, and a user-friendly interface.