摘要
本文提出一种新的方法实现森林火情自动检测。首先采用优化的码书模型模拟森林场景,实现森林火情发生区域的自动检测。然后对检测出的二值火情区域进行中值滤波去噪,消除背景干扰。最后根据列像素和的统计情况,设定阈值自动判断是否发生火情。实验结果表明本文方法具有很高的准确率,对于森林火情的自动检测十分可行。
This paper presents a new method to realize automatic detection of forest fire.First,the optimized codebook model is used to simulate the forest scene to realize the automatic detection of forest fire occurrence area.Then median filter is used to denoise the detected binary fire area to eliminate background interference.Finally,according to the statistics of column pixel sum,the threshold value is set to automatically determine whether fire occurs.The experimental results show that this method has high accuracy and is very feasible for automatic detection of forest fire.
作者
李国良
Li Guoliang(Institute of Highland Forest Science,Chinese Academy of Forestry,Kunming Yunnan,650233)
出处
《电子测试》
2022年第24期60-62,共3页
Electronic Test
关键词
森林火情
码书模型
中值滤波
阈值检测
forest fire
Codebook model
Median filtering
Threshold detection