煤泥浮选泡沫数字图象处理研究(之二)——煤泥浮选泡沫视觉特征的面邻域提取算法  

Research on Digital Image Processing of Coal Flotation Froth(Ⅱ) --The square neighbor algorithm for extracting features of digital coal flotation froth image

作　　者：刘文礼[1] 路迈西[1] 王振翀[1] 王勇[1] 王凡[1] 

机构地区：[1]中国矿业大学化工与环境工程系,北京100083

出　　处：《中国矿业大学学报》2002年第3期233-236,共4页Journal of China University of Mining & Technology

基　　金：国家自然科学基金资助项目 ( 59974 0 3 2 ) ;博士点基金资助项目 ( 1 9990 2 90 1 1 )

摘　　要：提出了有效描述浮选泡沫视觉特征的面邻域提取算法——邻域灰度相关矩阵法 ;引入了基于该矩阵的细度参数、粗度参数、二阶距参数、熵参数及不均匀性等特征参数来描述泡沫的结构 .利用笔者开发的煤泥浮选泡沫特征参数提取软件完成了分批浮选精矿泡沫图象的特征参数提取工作 ,并定性地分析了各泡沫特征参数与泡沫纹理的相关性 .结果发现细度参数、粗度参数能较好地反映出浮选泡沫图象的纹理特征 ,但熵参数。The square neighbor algorithm--neighboring gray level dependence matrix (NGLDM) for describing the visual textural features of flotation froth was put forward. Parameters, based on the NGLDM, such as fineness, coarseness, entropy, second moment, and non uniformity parameter were introduced to describe the texture of froth. The software developed by the authors was used to perform, in batches, the extraction of characteristic parameters of the froth image of flotation clean coal and to analyze qualitatively the relativity of characteristic parameter to texture of froth. The result shows that the parameters of fineness and coarseness may well reflect the textural characteristic of flotation froth images, while the parameters of entropy, second moment, and non uniformity have only a little relativity to the textural characteristic.

关 键 词：视觉特征 面邻域提取算法 煤泥 浮选泡沫 数字图象处理 

分 类 号：TD94[矿业工程—选矿] TP391.41[自动化与计算机技术—计算机应用技术]