基于泡沫图像特征融合的煤泥浮选工况识别  

Coal slime Flotation Condition Identification Based on Fusion of Froth Image Features

作　　者：梁秀满 田童 刘文涛 牛福生 LIANG Xiu-man;TIAN Tong;LIU Wen-tao;NIU Fu-sheng(College.of Electrical Engineering,North China University of Science and Technology,Tangshan Hebei 0632101,China;College of Mining Engineering,North China University of Science and Technology,Tangshan Hebei 063210,China)

机构地区：[1]华北理工大学电气工程学院,河北唐山063210 [2]华北理工大学矿业工程学院,河北唐山063210

出　　处：《计算机仿真》2021年第4期385-389,共5页Computer Simulation

基　　金：国家自然科学基金资助项目(51874135)。

摘　　要：煤泥浮选泡沫的图像特征与浮选工况密切相关。为提高图像特征分类的准确率,提出一种泡沫亮点分布特征、图像灰度特征与Tamura纹理特征相融合的方法用以表征浮选工况状态。首先利用Otsu阈值分割与形态学开操作提取泡沫的亮点,统计其大小、数量等分布特征;然后根据图像灰度值计算像素均值、方差和峰度方差;再提取泡沫图像Tamura纹理的粗糙度、对比度与方向度参数。将以上三类图像特征线性排列,标记特征数据建立支持向量机分类模型,并验证分类效果。结果表明,所提出的多种图像特征融合的方法的分类准确率达到88.6%,优于单一特征的分类结果。The image characteristics of slime flotation froth are closely related to flotation conditions. In order to improve the accuracy of image feature classification, a fusion method combining the features of froth highlights distribution, image grayscale features and Tamura texture features was proposed to characterize the state of flotation conditions. First, we used Otsu threshold segmentation and morphological opening operation to extract the highlights of the froth, count the features of size, quantity, etc. Then according to the image gray value, the pixel mean, variance and kurtosis-variance were calculated. And then the coarseness, contrast and directionality parameters of Tamura texture were extracted. The features of the above three categories of image were arranged linearly, and the Support Vector Machine(SVM) classification model was established with the feature data, and the classification effect was verified. The results show that the classification accuracy of the proposed method is 88.6%,which is better than that of single feature.

关 键 词：煤泥浮选 亮点分布 灰度特征 纹理特征 支持向量机分类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]