A Real-Time Fire Detection Method from Video with Multifeature Fusion

Abstract

The threat to people’s lives and property posed by fires has become increasingly serious. To address the problem of a high false alarm rate in traditional fire detection, an innovative detection method based on multifeature fusion of flame is proposed. First, we combined the motion detection and color detection of the flame as the fire preprocessing stage. This method saves a lot of computation time in screening the fire candidate pixels. Second, although the flame is irregular, it has a certain similarity in the sequence of the image. According to this feature, a novel algorithm of flame centroid stabilization based on spatiotemporal relation is proposed, and we calculated the centroid of the flame region of each frame of the image and added the temporal information to obtain the spatiotemporal information of the flame centroid. Then, we extracted features including spatial variability, shape variability, and area variability of the flame to improve the accuracy of recognition. Finally, we used support vector machine for training, completed the analysis of candidate fire images, and achieved automatic fire monitoring. Experimental results showed that the proposed method could improve the accuracy and reduce the false alarm rate compared with a state-of-the-art technique. The method can be applied to real-time camera monitoring systems, such as home security, forest fire alarms, and commercial monitoring.

HelpResponder—System for the Security of First Responder Interventions

“…The objective is to rule out regions that have passed the color filter but are not strong enough to be fire. The boundaries disorder is a parameter to analyze objects discarding objects with a regular shape of [37], centroids, and contours.…”

Fire Detection Methods Based on Various Color Spaces and Gaussian Mixture Models

“…( , ) > 190; ( , ) > 90; ( , ) < 140 (2) However, fire-like colors are sometimes misclassified as fire, leading to false alarms. As mentioned previously, both luminance (intensity) and chrominance changes in direct proportion to illumination.…”

“…Most of the work in fi re detection in color images and videos are based on setting color rules. [1,2] proposed a method of developing a set of rules for fi re pixel classifi cation based on video image processing. The RGB color system was used to extract fi re pixels and smoke pixels, which are then verifi ed if they are real fi re pixels or fi re-like pixels.…”

Frame Difference Method to Detect Fire and Compare the Accuracy and Precision with Vibe Method

“…It has to be controlled before it destroys the natural resources. Many incidents happened and many people have died, some are injured and 3.72 acres of the forest destroyed (Gong et al 2019) ; (Boesak and Hansen 2009). By capturing the video and dividing the video into frames (Yaoming et al 2017).Each and every frame is monitored and then noticed that, whether the fire is presented or not.…”