摘要
相比较传统以温度、烟雾等传感器为主的火灾检测方法,通过对火焰的成像特征进行分析,采用视频检测火灾。针对可见光下的火焰特征,基于计算机视觉技术,设计并实现了一种基于颜色模型的火灾识别系统。本系统以火灾检测算法为核心,设计并实现了简洁、易用的可视化的界面,将多种图像处理算法融入其中,并通过实验数据测试表明该系统的火灾识别平均准确度能够达到96%以上,与传统算法相比具有更低的误报率和更高的识别灵敏度。
Compared with the traditional fire alarm system based on temperature,smoke and other sensors,video fire detection through the analysis of the flame imaging features,according to the flame features under visible light,a fire recognition system based on color model is designed and implemented based on computer vision technology.This system takes the fire detection algorithm as the core,designs and implements a simple and easy-to-use visual interface,integrates a variety of image processing algorithms into it.A large number of experimental data tests show that the average accuracy of the system can reach more than 96%,which has lower false alarm rate and higher recognition sensitivity compared with the traditional algorithm.
作者
刘洲岐
王雷
刘聪
黄晋
王振
LIU Zhouqi;WANG Lei;LIU Cong;HUANG Jin;WANG Zhen(School of Computer Science and Technology,Shandong University of Technology,Zibo 255049,China)
出处
《山东理工大学学报（自然科学版）》
CAS
2022年第3期1-6,共6页
Journal of Shandong University of Technology:Natural Science Edition
基金
国家自然科学基金项目(61902222)
山东省泰山学者工程专项基金项目(tsqn201909109)。
关键词
视频监控
火焰识别
计算机视觉
特征提取
高斯混合模型
video monitoring
flame recognition
computer vision
feature extraction
Gaussian mixture model
