Research on fire warning system based on image feature extraction and multisensor fusion

Abstract

With the advancement of image processing technology, vision based fire detection has gradually become a research hotspot. This paper proposes a fire warning system design based on image feature extraction and multi-sensor fusion. The system is based on FPGA and captures images through OV7725 cameras. It uses HSV color space transformation and morphological processing algorithms to extract fire features, and integrates smoke and temperature sensor data for comprehensive judgment. Through experimental testing, the system can effectively remove environmental noise, calculate the center coordinates of fire targets in real time, and drive servo tracking, while rendering and displaying annotation information on the screen. The multi-sensor fusion mechanism significantly reduces the false alarm rate of a single sensor, and this system will be widely applied in fields such as intelligent security and industrial monitoring.