The swift increase of fire-related accidents in residential and industrial areas demands efficient and automated safety techniques. Traditional fire safety systems, which are based on passive smoke and temperature sensing, have shown limited performance in keeping up with fire safety demands. These systems have reputations for poor promptness, too many false alarms, and indiscriminate suppression mechanisms, leading to extensive collateral damage. The current paper aims to reduce these challenges with a new, cost-effective, and efficient Automated Fire Suppression System (AFSS) based on Internet of Things (IoT) and Computer Vision techniques. The proposed method makes use of ESP32-CAM, which can efficiently capture video data. Computer Vision is then applied to efficiently recognize fire using the HSV color range segmentation method. The method then makes use of a PI control mechanism to accurately locate a water nozzle to align with the center of a fire, ensuring effective fire suppression. The water flow is then initiated using a water pump. The proposed method is tested with efficient results, obtaining 94% accuracy in fire detection for different illumination conditions. The method has a response time of less than 2 seconds, ensuring efficiency in fire suppression.
Prasad et al. (Thu,) studied this question.