基于火焰图像红外动态特征的早期火灾识别算法  

Algorithm for Early Fire Recognition Based on Infrared Dynamic Characteristics of Flame Images

作　　者：王思嘉[1] 裴海龙[1] 

机构地区：[1]华南理工大学自动化学院,广东广州510640

出　　处：《现代电子技术》2010年第8期104-105,110,共3页Modern Electronics Technique

摘　　要：针对普通彩色图像对早期火灾高危点的有限检测能力,提出一套基于红外图像动态特征的早期火灾识别算法,该算法使用改进的梯度算子结合图像的动态特征来判断火焰。通过实验,该算法能够快速地识别火灾火焰和早期火灾高危点。算法结构清晰,效率较高,便于工程实现。Since the detection ability of ordinary color images for early fire high-risk points is limited, the early fire recognition algorithm based on infrared image dynamic characteristics is proposed. The algorithm judges the harmfulness of the flame by the modified gradient operator in combination with the dynamic characteristics of the images, and can quickly identify fire flame and high-risk points of early fire. The experiment proves that the algorithm structure is clear, efficient, and easy to be implemented in engineering.

关 键 词：红外图像 火灾火焰 动态特征 识别算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]