Video Fire Detection Algorithm using Multi-Feature Fusion

At present, the moving target detection and flame characteristics extraction almost become the most important parts in majority of video fire detection systems. Through the above two-part study, a new fire features detection method is presented in precise moving target area. That is, using the improved background difference method and flame features (such as the color and uniformity, Wavelet energy, stroboscopic and contour features) to detect fire. Experiments show that this method can improve the accuracy and anti-interference ability of fire detection. DOI: http://dx.doi.org/10.11591/telkomnika.v11i10.3340