考虑图像颜色小波纹理特征的工业炉温度检测关键技术  

Key Technologies for Industrial Furnace Temperature Detection Considering Wavelet Texture Features of Image Color

作　　者：杨超[1] YANG Chao(Yantai Vocational College,Yantai 264670,China)

机构地区：[1]烟台职业学院智能制造系,山东烟台264670

出　　处：《工业加热》2024年第9期57-61,共5页Industrial Heating

摘　　要：工业炉温度检测对工业炉运行状态分析具有重要意义,因此提出考虑图像颜色小波纹理特征的工业炉温度检测方法。首先,结合Retinex算法和TopHat变换对工业炉图像展开增强处理,提高特征提取精度;其次,在图像颜色空间变换的基础上获取工业炉图像的颜色特征,在小波变换的基础上获取工业炉图像的颜色小波纹理特征;最后,在最小二乘支持向量机中输入颜色小波纹理特征,实现工业炉温度的检测。实验结果表明,所提方法的图像增强效果显著、温度检测精度高,且温度检测时间控制在0.3 s以内。有效提高工业炉温度检测效果,具有较高的实用性。The temperature detection of industrial furnaces is of great significance for the analysis of their operating status.Therefore,a temperature detection method for industrial furnaces is proposed that considers the wavelet texture features of image color.Firstly,combining Retinex algorithm and TopHat transform to enhance the industrial furnace image and improve the accuracy of feature extraction;Secondly,based on the color space transformation of the image,the color features of the industrial furnace image are obtained,and the color wavelet texture features of the industrial furnace image are obtained on the basis of wavelet transform;Finally,input color wavelet texture features into the least squares support vector machine to achieve temperature detection of industrial furnaces.The experimental results show that the proposed method has significant image enhancement effect,high temperature detection accuracy,and a temperature detection time controlled within O.3 seconds.Effectively improving the temperature detection effect of industrial furnaces has high practicality.

关 键 词：TopHat变换 RETINEX算法 小波变换 工业炉温度检测 最小二乘支持向量机 

分 类 号：TM924[电气工程—电力电子与电力传动]