首页|基于视觉定位的磁铁矿分选方案设计

基于视觉定位的磁铁矿分选方案设计

扫码查看
针对某企业磁铁矿预选分选设备存在矿石定位不准和矿石边缘位置重叠难以实现连续分选的问题,设计了 一种基于视觉定位的磁铁矿预选分选方案.该方案采用矿石磁信号阈值分割、视觉定位、分水岭算法分割得到磁铁矿类别及形心位置,并设计分选控制策略,从而实现合格矿分选.试验结果表明,该方案可实现磁铁矿准确定位与边缘重叠分割,并连续分选合格矿,平均选别效率为53.83%;单张矿石图像的平均处理时间为0.216 s,单个矿石颗粒的分选时间约为20 ms,可满足快速分选矿石的需求.
Design of Magnetite Separation Scheme Based on Visual Positioning
There are inaccurate positioning of ore and overlapping of ore edge positions in the magnetite pre-concentration and separation equipment of an enterprise,resulting in the difficulty of continuous separation.Aiming at theses problems,a magnetite pre-concentration and separation scheme based on visual positioning was designed.In this scheme,ore magnetic signal threshold segmentation,visual positioning,watershed algorithm segmentation were used to obtain magnetite category and centroid position,and separation control strategy was designed to realize qualified ore separation.The results show that the scheme can realize accurate positioning and edge overlapping segmentation of magnetite,and continuously separate qualified ore,with an average separation efficiency of 53.83%.The average processing time of a single ore image is 0.216 s,and the separation time of a single ore particle is about 20 ms,which can meet the needs of rapid separation of ore.

Magnetite separationVisual positioningEdge segmentationMagnetic signal detection

利健、潘春荣、熊文清

展开 >

江西理工大学机电工程学院,江西赣州市 341400

磁铁矿分选 视觉定位 边缘分割 磁信号检测

江西省研究生创新专项项目

YC2021-S577

2024

矿业研究与开发
长沙矿山研究院有限责任公司 中国有色金属学会

矿业研究与开发

CSTPCD北大核心
影响因子:0.763
ISSN:1005-2763
年,卷(期):2024.44(8)