摘要
针对基于PC机的辐射测温系统存在结构复杂、体积较大、不便携带和安装使用等问题,提出了基于DSP的高温辐射体表面温度场快速测量方法。结合多光谱图像分割技术与改进的一维快速Otsu阈值算法,实现高温目标图像的快速识别,然后以高温辐射体表面图像像素点的红、绿基色值作为索引,运用查表法快速获取高温辐射体表面温度场分布,并以伪彩色图像实时显示。实验结果表明:该方法能够在基于DSP搭建的硬件平台上快速有效地测量出高温辐射体表面温度场分布,具有较强的实用性。
Considering radiation thermometry systems based on PC,which are structural complex and inconvenient to carry and assemble,a rapid method of measuring facial temperature field of high-temperature radiators based on DSP was put forward. This method combined segmentation of multispectral image with advanced one dimension Otsu threshold method recognized the image of high-temperature radiators rapidly,and used red and green color values of that image＇s pixels as index to acquire the facial temperature field distribution of high-temperature radiators in terms of table look-up. The facial temperature field was shown as pseudo color image in real time. Experiments show that this method can measure facial temperature field of high-temperature radiators on DSP hardware rapidly and efficiently,which has wide applicability.
出处
《仪表技术与传感器》
CSCD
北大核心
2015年第12期145-147,共3页
Instrument Technique and Sensor
基金
国家自然科学基金面上项目(61273169)
湖南省教育厅高等学校科研项目(13A016)
关键词
辐射测温
快速图像识别
比色测温
嵌入式系统
radiation thermometry
fast image recognition
look up table
embedded system
