嵌入式智能图像处理在火灾预警系统的应用

传统的火灾检测方法一般采用感烟、感温、感光探测器等进行探测。本文提出了一种嵌入式基于图像视觉特征的火灾检测方法,以TI公司的数字多媒体处理器TMS320DM642为核心,设计实现智能前端火灾探测与自动报警系统。通过DM642对视频图像进行采集并结合相应的智能图像处理与模式识别算法,对森林火险进行实时监控。实验结果表明,该系统比传统系统更进一步减少了误报率且具有响应快、监控范围广等优点。

Traditional forest fire detection methods generally use smoke,temperature,photosensitive detector. This paper presents a visual image based on the embedded characteristics of the fire detection methods, TMS320DM642 digital media processor as the core from TI Company,intelligent front-end design to achieve automatic fire detection and alarm system.DM642 capture the image Sequences and make use of the smart image processing and pattern recognition algorithms,monitoring the forest fire on real-time.Experimenta...