摘要
船舶机舱火灾检测是个极为重要的环节。利用嵌入式系统的优点结合机器视觉技术成为这个领域的一个研究方向。机器视觉的方法被逐步应用于火灾检测中,而火灾视觉特征的提取是主要研究方向。本文以火灾的轮廓脉动特征为主要研究点,通过对轮廓的变化描述来探测火灾。利用离散余弦变换对轮廓信息的描述,同时在离散余弦变化空间上构筑脉动信息的距离模型。通过该模型准确地度量时空闪烁特性,从而对火灾进行判断。试验表明,本方法具有较好的鲁棒性和准确性,并可较好地提高系统的运行速度,减少运算花费的时间。
The ship engine-room fire detection is very important, so the research on it is regarded. Using the strongpoint of embedded system with machine vision will be a research point in this field. The method of machine vision is gradually used in fire detection, in which the pick-up of fire vision characteristic is a main research field. This paper＇s main research point is the characteristic of outline fluctuation. According to description of outline change,we can detect the fire. Using discrete cosine transformation to descript the information of outline. The model of contour fluctuation can be builded. Using this model to detect outline fluctuation feature of flame,we can detect the fire better. Experiments show that the algorithm is effective and robust, and that it is significant for improving accuracy and increasing the speed of system.
出处
《舰船科学技术》
北大核心
2008年第5期54-58,共5页
Ship Science and Technology
基金
船舶行业国防预研项目(03J3.61)
关键词
机器视觉
船舶机舱火灾探测
嵌入式系统
离散余弦变换
machine vision
ship engine-room fire detection
embedded system
discrete cosine transformation
