No abstract

Abskuct-Anew image-based real-time flame detection method is proposed in this paper. First, fire flame features based on the HS1 color model are extracted by analyzing 70 flame images. Then, based on these flame features, regions with fire-like colors are roughly separated from an image. Besides segmenting fire flame regions, background objects with similar fire colors or caused by color shift resulted from the reflection of lire flames are also separated from the image. In order to get rid of these spurious lire-like regions, the image difference method and the invented color masking technique are applied. Finally, a simple method is devised to estimate the burning degree of fire flames so that users could be informed with a proper warning alarm. The proposed method is tested with seven diverse fire flame video clips on a Pentium I1 350 processor with 128 MB RAM at the process speed of thirty frames per second. The experimental results are quite encouraging. The proposed method can achieve more than 96.97% detection rate on average. In addition, the system can correctly recognize fire flirmes within one second on the initial combustion from the test video clips, which seems very promising.Index Terms--Color analysis, color masking, burning degree estimation, flame detection.

No abstract

An image-based fire detection method using neural networks is proposed in this paper. First, flame color features, based on the HSI color model, are trained by a backpropagation neural network for flame recognition. Then, based on the learned flame color features, regions with fire-like colors are roughly separated from an image. Besides segmenting flame regions, background objects with similar fire colors or resulted from the reflection of fire flames are also separated from the image. In order to get rid of these spurious fire-like regions, the image difference method and the invented color masking technique are applied. Finally, a compact method is devised to estimate the burning degree of fire flames so that users could be informed with a proper warning alarm. The proposed system can achieve 96.47% fire detection rate on average.

No abstract