烧结矿矿相灰度直方图中矿物信息的提取与分析  

Extraction and analysis of mineral information in gray histogram of sinter ore mineragraphy

作　　者：邱贵宝[1] 吕学伟[1] 白晨光[1] 黄青云[1] 刘松利[1] 牛文娟[1] 

机构地区：[1]重庆大学材料科学与工程学院,重庆400044

出　　处：《冶金分析》2008年第8期7-11,共5页Metallurgical Analysis

基　　金："973计划"资助项目(2007CB13503);重庆大学研究生科技创新基金项目(200701Y1B0020188)

摘　　要：智能矿相识别处理系统将会为矿物鉴定及测定工作带来便利,准确和全面的矿物图像特征提取是关键。本文运用矿物反射率计算模型,推导得出烧结矿常见物相在灰度直方图曲线上"峰"的相对位置分布关系。针对矿物显微图像,得到其灰度直方图后,利用曲线寻峰模型获取灰度直方图"峰"数及"峰"位信息,并以此作为鉴别矿物种类和组成的依据。经实验验证,反射率计算模型和曲线寻峰模型作为图像特征提取方法,能够获得矿物种类信息,为进一步实现智能化矿相识别打下基础。The intelligent recognizing and processing system for sinter ore mineragraphy gives a great convenience for the recognition and measurement of minerals. Specially, it＇s critical important for a successful intelligent system to extract the minerals＇ feature accurately and sufficiently. The paper got the relationships of different minerals＇ peak-positions in curve of gray histogram by using the model of index of reflection. For a mineral micrograph, the information of peak number and position, which are regarded as the references to identify mineral categories and contents, can be got by using model of curve peak-searching after gray histogram is got. The models of index of reflection and peaksearching can get the mineral information, which lays a good foundation for the artificial intelligent system of recognition and processing in sinter ore.

关 键 词：烧结矿矿相 反射率 曲线寻峰 灰度直方图 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]