基于改进FCM和形态学的浮选泡沫形态特征提取

来源期刊:中南大学学报(自然科学版)2010年第3期

论文作者:周开军 王一军 许灿辉

文章页码:994 - 1000

关键词:浮选;泡沫图像;形态特征;模糊C均值;数学形态学

Key words:flotation; froth image; morphological feature; fuzzy C-means; mathematic morphology

摘    要:针对浮选过程中因气泡粘连及形状不规则导致泡沫形态特征难以提取的问题,提出一种基于改进模糊C均值(fuzzy C-means, FCM)聚类和数学形态学的浮选泡沫形态特征提取方法。引入聚类有效性指数及特征散度对模糊C均值聚类算法加以改进,并利用改进的聚类算法对泡沫图像进行聚类,得到泡沫大致区域。依据灰度分布和形状特征,采用面积重构开闭算法对图像进行除噪处理。基于形态重构方法思想,提出采用高低精度距离变换方法,同时,结合改进面积重构变换提取标志图像,进而利用分水岭算法对泡沫图像进行分割。通过测量分割区域和标定像素提取泡沫形态特征,并与浮选工艺参数做相关性分析。研究结果表明,该方法能够准确地分割粘连泡沫,且提取的泡沫形态特征能有效反映浮选工况。

Abstract: Considering the difficulty of morphological feature extraction that resulted from bubble adhesion and irregular shape in flotation process, a novel froth morphological feature extraction method based on fuzzy C-means (FCM) clustering and mathematic morphology was presented. FCM algorithm was improved by introducing the clustering validity index and feature divergence for the real-time performance. In order to obtain the rough region of bubble, the froth image was clustered using the improved clustering algorithm, and the noises were filtered using area reconstruction algorithm in terms of intensity and shape distribution information. Based on morphological reconstruction, a high-low scale distance transformation was brought forward, with which the region markers were extracted by combining the area reconstruction. Furthermore, the froth image was segmented by watershed algorithm. The morphological features were extracted by measuring the segmented regions and demarcating the pixels, and some correlation analyses between the froth characters and flotation technical parameters were brought forward. The experimental results show that the proposed method can segment adherence froth image exactly, whilst the morphological features can reflect floatation operating condition effectively.

基金信息:国家自然科学重点基金资助项目
国家自然科学基金资助项目

相关论文

  • 暂无!

相关知识点

  • 暂无!

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号