基于LBPV的浮选泡沫图像纹理特征提取

摘要

在浮选生产中,浮选泡沫表面纹理与浮选工况密切相关,直接反映泡沫层的矿化程度(品位高低).为了给浮选操作提供指导,提出了一种基于LBPV( local binary pattern variance)的泡沫图像纹理特征提取方法.该方法通过融合泡沫图像局部空间结构和对比度来提取泡沫图像纹理特征,然后将LBPV纹理特征应用于浮选工况状态的聚类分析.结果表明,该方法提取的纹理特征能有效反映浮选工况,且能获得更优的浮选泡沫聚类质量.%In the flotation process, the surface features of the cleaner froth had close connection with the flotation production condition and reflected the mineralization of foam layer( concentrate grade). In order to optimize the flotation operation, this paper proposed a method of LBPV to extract the bubble texture features. The features of bubble image were extracted on the combination of local texture spatial structures and contrast features. LBPV texture features used in the cluster analysis of the flotation operation condition state. Clustering results show that the texture feature are efficient in reflecting the bubble states and can obtain better clustering quality of flotation foam.