Enhanced Fire and Smoke Detection Utilizing Yolov8: A Deep Learning Approach

Fire and smoke are hazardous environmental phenomena that pose significant risks to safety and security, necessitating effective detection and monitoring systems. In this study, a new framework based on the YOLOv8 object detection paradigm is introduced for unified fire and smoke identification in pictures and streaming video streams. Our method achieves outstanding accuracy with a mean average precision (mAP) of 98.5% by utilizing the sophisticated features of YOLOv8. Evaluation metrics include F1 score, precision, and recall, ensuring comprehensive assessment. We describe the creation of a YOLOv8-based smoke and fire detecting system, with an emphasis on precise and effective instance recognition. The project involves dataset preparation, model training, and evaluation using essential metrics. Furthermore, the project emphasizes the development of a Graphical User Interface (GUI) for prediction visualization. The GUI enhances the user experience by providing a user-friendly interface to interact with the detection system. Users can input images or video streams, and the GUI will display the predictions, including bounding boxes around detected fire and smoke regions.