Fire and Smoke Detection from IP Camera Footage using Faster R-CNN with ResNet101V2

Fire and Smoke detection is a critical component of fire safety systems, enabling early detection and mitigation of fire incidents. Traditional smoke detection methods often rely on specialized hardware, which have bieng costly and challenging to deploy in large-scale environments. The proposed paper come up with a new idea to detect smoke using deep-learning and neural networks, specifically leveraging the Faster R-CNN architecture with a ResNet101V2 backbone for analyzing video streams from IP cameras. The system preprocesses the video footage to enhance quality and remove noise, ensuring accurate detection. The Faster R-CNN with ResNet101V2 backbone architecture is utilized for its performance in object detection tasks, allowing us to detect smoke with accuracy and efficiency.