摘要
针对煤泥浮选泡沫图像噪声大、黏连性强的问题,提出了一种煤泥浮选泡沫图像的分割方法:首先,增强煤泥浮选泡沫图像的对比度,以提高后续煤泥浮选泡沫图像的识别度;然后,采用不断增大结构元素的面积,重构开闭运算滤除不同的噪声,通过面积重构H顶变换和模糊C均值聚类(Fuzzyc-means clustering,FCM)提取泡沫图像的标识;最后,采用分水岭变换技术分割泡沫图像。该研究为煤泥浮选泡沫提供一种图像分割方法,同时为后续煤泥浮选自动化作业打下基础。
In view of huge noise and severe combination of coal slurry flotation froth image, a segmentation method for coal slurry flotation froth image was proposed. At first, the contrast of the coal slurry flotation froth image increased so as to enhance the recognition of the subsequent coal slurry flotation froth image. And then, the area of structuring elements was increased unceasingly as well as opening and closing operation was reconstructed so as to filter various noises, and area reconstruction H-dome transformation and fuzzy c-means clustering (FCM) were applied to extract the marks of the froth image. Finally, the watershed transformation technology was used to segment the froth image. The method provided another segmentation alternative for coal slurry flotation froth image, and laid foundation for subsequent automated operation of coal slurry flotation.
出处
《矿山机械》
北大核心
2013年第3期103-106,共4页
Mining & Processing Equipment
基金
山西省科技攻关项目(20120321004-32)
关键词
面积重构
模糊C均值聚类
分水岭变换
area reconstruction
fuzzy c-means clustering
watershed transformation