基于图像识别的井下机车载矿量计量系统  

Mine-Loading Measurement System of Underground Locomotive Based on Image Recognition

作　　者：张钰鹏 吴富姬 郭毅 ZHANG Yupeng;WU Fuji;GUO Yi(Nonferrous Metal Mining and Metallurgy Equipment Work Design Center,Ganzhou Nonferrous Metallurgy Research Institute Co.,Ltd.,Ganzhou 341000,Jiangxi,China)

机构地区：[1]赣州有色冶金研究所有限公司,有色金属矿冶装备工作设计中心,江西赣州341000

出　　处：《黄金科学技术》2022年第1期131-140,共10页Gold Science and Technology

基　　金：江西省重点研发计划项目“基于图像识别和‘大数据’的非接触式运矿计量系统成套技术及装备研发”(编号:20202BBEL53016)资助。

摘　　要：出矿量是矿山企业制定生产计划的一个重要指标,而目前绝大多数地下矿山采用人工清点载矿矿车数的方式估算出矿量,该计量方式存在估算误差大等缺点,严重影响了矿山企业生产计划的合理制定。本文采用图像识别与密度模型建模相结合的方法,通过对矿石堆图像信息进行三维重建形成体积模型,构建了一个图像特征—密度库,设计并开发了一套基于图像识别的井下机车载矿量计量系统。经现场反复试验证明,该计量系统运行稳定可靠,机车载矿量计算误差小于5%,较好地解决了目前矿山企业出矿量估算不准确的问题,提高了出矿量计算的准确度,为矿山企业制定生产计划提供了详实可靠的数据。The ore yield is an important standard for mining enterprises to formulate production plans.At present,the vast majority of underground mines estimate ore yield by manually counting the number of ore carrying vehicles.The measurement error is large,which seriously affects the enterprises to formulate production plans.In order to solve the problem of large measurement error and improve the accuracy of ore yield estimation,a set of underground machine on-board ore yield measurement system based on image recognition was designed and developed in this paper.In this paper,the method of image recognition combined with density model modeling was used to form a volume model through three-dimensional reconstruction of ore pile image information,and an image feature density library was built to form a complete set of underground machine onboard ore quantity measurement system.The system collects the internal image of the locomotive ore bucket through the depth camera,then extracts the feature information of the image and compares it with the image feature density library to obtain the density of the ore in the current ore bucket,and then generates a volume model from the three-dimensional reconstruction of the image to calculate the volume of the ore pile,and calculates the product of the volume and density of the ore pile to obtain the weight of the ore pile.The field repeated tests show that the metering system operates stably and reliably,and the calculation error of locomotive ore load is less than 5%.It solves the problem of ore yield estimation in mining enterprises,improves the calculation accuracy of ore yield,and brings detailed and reliable data for enterprises to formulate production plans.

关 键 词：图像识别 三维重建 密度模型 计量系统 井下机车 出矿量 

分 类 号：TP274.5[自动化与计算机技术—检测技术与自动化装置]