摘要
在总结国内外示温漆自动判读算法研究现状的基础上,详细讨论现有算法的优缺点,进一步研究示温漆的自动判读技术。通过引用红外特征的方法,并采用图像融合将其与示温漆颜色特征组合成新的图像数据,然后提取出新的图像数据的统计特征,采用机器学习算法自动建立等温线和像素点温度识别模型,两种模型相结合用于示温漆温度自动判读,达到提高示温漆测温准确度的目的。
The advantages and disadvantages of existing algorithms for automatic temperature reading were discussed and the automatic temperature reorganization techniques of thermal paint were researched according to the summarization of current research at home and abroad on automatic temperature recognition algorithm of thermal paint. Infrared feature was introduced in combination with color feature to form new image data by image fusion. Afterwards, statistical features were extracted from these new data, and two temperature identification models for isotherm line and pixel were automatically created by machine-learning algorithm. The two models were used to automatically identify the thermal paint to improve its temperature-measuring accuracy.
出处
《中国测试》
CAS
北大核心
2015年第9期20-23,共4页
China Measurement & Test
关键词
示温漆
自动判读算法
颜色温度特性曲线
颜色空间
thermal paint
automatic recognition algorithm
color-temperature curve
color space