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

Area of Science:

• Computer Vision
• Artificial Intelligence
• Fire Safety Engineering

Background:

• Traditional fire detection systems suffer from high false alarm rates.
• Accurate and timely fire detection is critical for protecting lives and property.

Purpose of the Study:

• To develop an innovative fire detection method using multifeature fusion of flame characteristics.
• To reduce the false alarm rate and improve the accuracy of fire detection systems.

Main Methods:

• Combined motion and color detection for fire preprocessing, optimizing computational efficiency.
• Developed a novel flame centroid stabilization algorithm using spatiotemporal relations for enhanced feature extraction.
• Extracted spatial variability, shape variability, and area variability of the flame.
• Utilized support vector machine (SVM) for training and automatic fire monitoring.

Main Results:

• The proposed multifeature fusion method demonstrated improved accuracy in fire recognition.
• Significantly reduced the false alarm rate compared to state-of-the-art techniques.
• Successfully achieved automatic fire monitoring through image analysis.

Conclusions:

• The innovative flame detection method offers a more accurate and reliable solution for fire monitoring.
• The technique is suitable for real-time camera monitoring systems, including home security, forest fire alarms, and commercial applications.