基于SMO算法的皮带撕裂红外图像检测方法  

SMO algorithm based infrared image detection method for belt tearing

作　　者：徐善永[1] 黄友锐[1] 冯涛 韩涛[1] XU Shanyong;HUANG Yourui;FENG Tao;HAN Tao(School of Electrical and Information Engineering,Anhui University of Science and Technology,Huainan 232001,China)

机构地区：[1]安徽理工大学电气与信息工程学院,安徽淮南232001

出　　处：《现代电子技术》2020年第11期37-40,46,共5页Modern Electronics Technique

基　　金：安徽省科技攻关计划(1501021027);安徽理工大学青年教师科研基金项目(QN201601)。

摘　　要：针对运煤皮带经常性的纵向撕裂问题,考虑到煤码头复杂环境引起的检测不精确性,提出基于序列最小最优化(SMO)算法的红外图像检测方法。由于煤码头存在着大量的水雾和粉尘,将在很大程度上影响图像的提取和处理。通过获取运煤皮带的红外图像,采用SMO算法构建决策模型并对红外图像进行分割。由实验效果图可得,分割效果良好,辨识度高,并从检测精度和分割时间两个角度出发,通过对比BP神经网络算法、SVM算法和SMO算法,表明SMO算法不仅预测精度高,而且实时性好,能够满足皮带撕裂图像检测的诊断要求。In view of the frequent longitudinal tearing of coal conveyor belts,an infrared image detection method based on sequential minimal optimization(SMO)algorithm is proposed by taking account of the detection inaccuracy caused by complex environment of coal terminal. There is a large amount of water mist and dust in the coal terminal,which will greatly affect the image extraction and processing. Therefore,by acquiring the infrared image of the coal belt,the decision-making model is built and the infrared image is segmented with the SMO algorithm. It can be seen from the experiment effect pictures that the segmentation results are good and the recognition degree is high. In addition,the BP neural network algorithm,SVM(support vector machine)algorithm and SMO algorithm are compared in the aspects of detection accuracy and segmentation duration. It shows that the SMO algorithm has not only high prediction accuracy,but also a fine real-time performance. Therefore,it can meet the diagnostic requirements of image detection for coal conveyor belt tearing.

关 键 词：红外图像检测 运煤皮带 纵向撕裂 SMO算法 图像分割 决策模型 

分 类 号：TN911.73-34[电子电信—通信与信息系统] TD634[电子电信—信息与通信工程]