The Design of the Flame Monitoring System Based on Machine-Vision with the Color Processing and Interframe Difference Method

Abstract

With the increasing emphasis on the safety, fire disaster caused by the uncontrollable flame is still a major factor that seriously affects the public safety and social development. In order to reduce the probability of fire disasters, flame monitoring systems are indispensable. But traditional flame monitoring systems are limited by their transducers. Aiming at these shortcomings, the flame monitoring system based on machine-vision with the color processing and interframe difference method is presented in this paper. The flame monitoring system with our method has been implemented via the FPGA platform, and the experimental results indicate that the flame can be distinguished correctly from other objects and environments, the maximum normalized color characteristics of the flame can also be provided for the flame monitoring in the large space.