摘要
传统的森林火灾监测技术包括人工护林监测、飞机航测、卫星监测等。本文提出一种基于普通CCD摄像头和瞭望塔上现有的短波无线通信设备的森林防火监测系统,根据实时图像与参考图像的差分及小波分解结果,当有异常情况发生时,提取出火焰及烟雾区域,判断提取区域是否具有烟雾和火焰的动态特征。若判定发生火灾,即将压缩后的图像通过短波通信设备传回指挥中心。
The traditional fire prevention technologies often involve manual monitoring, aerial survey and satellite monitoring, but they do not always work satisfactorily. A fire prevention system is proposed in this paper for detecting fire in the forest, based on the camera and the existing short-wave communication equipment on a tower. First, the results of the difference between the real-time image and the reference image and its wavelet decomposition are obtained. The fire and smoke regions are identified in case of abnormal situations. The segmentation method is used, based on color space, such as RGB and HSI. Then, the regions are decided by dynamic characters of fire and smoke. The image is compressed using JPEG2000 compression standard if fire is found in the region. Finally, the fire image is transmitted to the center room after compression through the short-wave communication equipment.
出处
《科技导报》
CAS
CSCD
2008年第5期77-80,共4页
Science & Technology Review
基金
黑龙江省科技攻关计划项目(2007G0757-00)
关键词
火灾识别
小波分解
图像提取
颜色聚类
fire recognition
wavelet decomposition
image extraction
color clustering
