| 摘要: |
| 森林火灾是一种突发性强、破坏性大且处置救助较为困难的自然灾害，会对森林、森林生态系统和人类带来一定危害和损失。早期的林火判别多基于温度、烟雾以及24小时图像监测实现，主观性强，标准单一，且误报率高。如何快速发现林火目标并及时报警扑灭，已经成为国内外学者研究的热点之一。文章基于红外热像仪，对大量森林背景及模拟林火状态进行了图像和温度数据的采集，首次提出了基于温度标准差方法进行热源识别的概念，得出了不同地点、不同时间段森林背景的温度标准差分布，以及明火温度在各阶段的波动范围，为林火的预防和监测提供了重要依据。同时，通过对干扰源进入前后的红外图像进行预处理和分析，对比目标热源的像素面积变化率及其圆形度，即可准确实现部分干扰热源的排除。实验结果证明，将此基于红外图像的热源初判原则与干扰源排除的规律应用于林火监测，方法简单有效，准确率高，对降低森林火灾的发生率，提高社会经济效益有着重要的作用。 |
| 关键词: 林火 红外图像 火源识别 温度标准差 预处理 |
| DOI：10.7621/cjarrp.1005-9121.20160803 |
| 分类号: |
| 基金项目: |
|
| IMPLEMENTATION OF FOREST FIRE RECOGNITION BASED ON INFRARED IMAGE |
|
Wu Zhaoming1, Zheng Change2, Shangguan Xiaorui3
|
|
1.Department of Electrical & Information Engineering，Nanjing Communications Institute of Technology，Nanjing， 211188，China;2.Beijing Forestry University，Beijing 100083，China;3.Chicago State University，Chicago IL 60628，America
|
| Abstract: |
| Forest fire is a kind of natural disaster which happens suddenly, had strongly destructive power, and difficult to handle with. It can bring huge harm and loss to the ecosystem and human production and development, and becomes one of the hottest research issues on how to find the forest fire timely, sound the alarm and extinguish the fire. Based on large amount of experiments about forest environment and forest fire by the infrared thermography, this paper firstly provided a method using standard deviation of temperature to distinguish heat source. Furthermore, it excluded some kind of interference heat sources according to their area changing ratio and circularity. The experiment results showed that the conclusions and rules to monitor forest fire were simple and effective, and can be applied to the practical environment with great accuracy. |
| Key words: forest fire infrared images fire source identification standard deviation of temperature pretreatment |
