基于数字图像处理的浮选泡沫速度特征提取及分析.docVIP

第 40 卷第 6 期 2009 年 12 月  中南大学学报(自然科学版) Journal of Central South University (Science and Technology)  Vol.40 No.6 Dec. 2009  基于数字图像处理的浮选泡沫速度特征提取及分析 唐朝晖，刘金平，桂卫华，阳春华 (中南大学信息科学与工程学院，湖南长沙，410083)  摘  要：以浮选泡沫图像序列为对象，研究浮选泡沫图像序列速度特征提取方法，分析泡沫速度特征与浮选性能 间的关系。提出一种抗尺度快速变化和具有旋转不变性的模板匹配算法，利用宏块跟踪技术对浮选泡沫图像序列 的泡沫速度特征进行估计，得出像素级的泡沫运动速度参量。然后，采用二维拉格朗日曲面插值方法提取亚像素 位移，得出精确的亚像素级位移参量。结果表明：在浮选过程中，减少浮选泡沫运动速度的紊乱程度能减小已粘 附在泡沫上的矿物粒子的脱附率，进而提高浮选精矿品位，降低尾矿中有用矿物含量。 关键词：浮选；泡沫图像；模板匹配；旋转不变性；亚像素配准 中图分类号：TP391.41 文献标志码：A 文章编号：1672?7207(2009)06?1616?07  Froth bubbles speed characteristic extraction and analysis based on digital image processing TANG Zhao-hui, LIU Jin-ping, GUI Wei-hua, YANG Chun-hua (School of Information Science and Engineering, Central South University, Changsha 410083, China) Abstract: Based on the study of the bubble image sequences, the characteristics of the forth velocities were extracted, and the relationship between the froth speed characteristic and the froth flotation performance was discussed. Hence, a new fast template matching algorithm was presented. This algorithm is invariant to arbitrary rotation revolving and scaling, and using macro-block tracking method to calculate the froth motion velocity, the integral pixels velocity characteristic can be obtained; then, the sub-pixels velocity parameters was extracted using the two-dimension Lagrange surface interpolation and the full precise sub-pixel scale froth bubbles motion velocity was gained. The results show that the decrease of froth velocity turbulence degree in the flotation production is able to decrease the detach ratio of the mineral particles on the bubbles and then to improve the concentrate grade and decrease the ratio of minerals in tailings. Key words: flotation; froth images; template matching; rational invariant; sub-pixels registration  浮选是一个极其复杂的物理化学过程，难以建立 准确的实时控制模型[1]。在实际的矿物浮选生产中， 有经验的操作工人主要通过观察浮选槽表面泡沫视觉 信息来完成浮选过程操作。浮选泡沫表面视觉特征参 数一般表现在 2 个方面：一是基于单帧图像的泡沫参 数，称为静态参数，比如泡沫的大小、形状、颜色特  征、纹理特征[2]等；另一方面是基于