基于机器视觉的矿石块度智能识别及应用  

Intelligent identification and application of oresize distribution based on machine vision

作　　者：朱颖舟 汪晓春 陆博 刘道喜 ZHU Yingzhou;WANG Xiaochun;LU Bo;LIU Daoxi(Anhui Tongguan(Lujiang)Mining Co.Ltd.,Hefei 238000,Anhui,China;BGRIMM Technology Group,Beijing 100160,China)

机构地区：[1]安徽铜冠(庐江)矿业有限公司,安徽合肥238000 [2]矿冶科技集团有限公司,北京100160

出　　处：《矿冶》2023年第4期119-125,共7页Mining And Metallurgy

基　　金：国家重点研发计划项目(2021YFC2902702)。

摘　　要：给矿粒度特性以及矿石可磨性对半自磨过程的影响远大于常规碎磨流程,通过对给矿皮带上矿石图像进行矿石粒度特性分析,可从给矿源头上实现矿石块度实时检测。基于多尺度的矿石图像分析系统,实时采集半自磨给矿矿石图像,通过图像处理技术,实现矿石块度有效分割以及粒级统计,为半自磨给矿矿石块度控制提供数据支撑。The impact of feed particle size characteristics and ore grind ability on the semi-automatic grinding process is much greater than that of conventional crushing and grinding processes.By analyzing the ore particle size characteristics of the ore images on the feed belt,real-time detection of ore block size can be achieved from the feed source.Based on a multi-scale ore image analysis system,real-time collection of semi-automatic grinding ore images is achieved.Through image processing technology,effective segmentation of ore block size and particle size statistics are achieved,providing data support for controlling the block size of semi-automatic grinding ore.

关 键 词：机器视角 半自磨机 矿石块度 图像分析 智能识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]