基于深度学习的安徽黄屯铜金矿床黄铜矿定量分析及其意义

doi:10.19657/j.geoscience.1000-8527.2024.077

现代地质 ›› 2025, Vol. 39 ›› Issue (03): 541-551.DOI: 10.19657/j.geoscience.1000-8527.2024.077

• 机器学习与矿物学应用 • 上一篇下一篇

基于深度学习的安徽黄屯铜金矿床黄铜矿定量分析及其意义

唐赧钰¹(), 申俊峰¹^,^*(), 陈强², 卿敏³, 赵永建⁴, 刘海明⁵, 董博², 陈满³, 龚智诚², 张萌萌¹, 林浩¹, 王浩阳¹

1.中国地质大学(北京)地球科学与资源学院成因矿物学研究中心,北京 100083
2.北京中育神州数据科技有限公司,北京 100043
3.比优探索(北京)资源勘查有限公司,北京 100010
4.安徽省金鼎矿业股份有限公司,安徽合肥 231555
5.中国地质科学院矿产资源研究所自然资源部成矿作用与资源评价重点实验室,北京 100037

出版日期:2025-06-10 发布日期:2025-07-03
通信作者: *申俊峰,男,教授,博士生导师,1962年出生,主要从事成因矿物学与找矿矿物学研究工作。Email:shenjf@cugb.edu.cn。
作者简介:唐赧钰,男,硕士研究生,2000年出生,主要从事成因矿物学与找矿矿物学研究工作。Email:2001220096@email.cugb.edu.cn。
基金资助:
教育部科技发展中心中国高校产学研创新基金项目(KBB036)

Quantitative Analysis of Chalcopyrite in Huangtun Copper-Gold Deposit in Anhui Province Based on Deep Learning and Its Significance

TANG Nanyu¹(), SHEN Junfeng¹^,^*(), CHEN Qiang², QING Min³, ZHAO Yongjian⁴, LIU Haiming⁵, DONG Bo², CHEN Man³, GONG Zhicheng², ZHANG Mengmeng¹, LIN Hao¹, WANG Haoyang¹

1. Research Center of Genetic Mineralogy, School of Earth Sciences and Resources, China University of Geosciences (Beijing), Beijing 100083, China
2. Beijing Zhongyu Shenzhou Data Technology Co., Ltd., Beijing 100043, China
3. Biyou Exploration (Beijing) Resources Exploration Co., Ltd., Beijing 100010, China
4. Anhui Jinding Mining Co., Ltd. Hefei, Anhui 231555, China
5. Ministry of Natural Resources (MNR) Key Laboratory of Metallogeny and Mineral Assessment, Institute of Mineral Resources, Chinese Academy of Geological Sciences (CAGS), Beijing 100037, China

Published:2025-06-10 Online:2025-07-03

摘要/Abstract

摘要：

长期以来,产业链上游或前端(找矿勘查)至中下游或后端(采选冶)始终采用有价元素作为定量指标。然而,由于有价元素赋存矿物的多样性(同一种元素可以赋存于多种矿物相态产出),采用元素定量往往造成产业前端的资源储量评价和后端的资源回收评价出现很大偏差。显然,由于选矿工艺过程主要是回收赋存可回收元素的目标矿物,采用目标矿物含量代替目标元素含量作为定量指标更加合理。基于上述原因,针对可回收目标矿物,采用人工智能图像识别技术,进行目标矿物快速准确的定量方法非常值得深入研究。安徽黄屯铜金矿床属于热液型矿床,其中Cu是主要回收资源之一,黄铜矿是Cu的主要赋存矿物相。选择矿区5条勘探线13个钻孔合计114件样品,采用显微图像深度学习方法,进行了黄铜矿识别与定量。结果显示,基于深度学习的人工智能图像识别技术可以准确识别黄铜矿并进行准确定量,而且黄铜矿矿物定量比Cu元素定量更能反映铜资源的空间变化规律和空间分布特点,对指导深部铜矿或相关矿种(金矿)找矿、矿床勘探、采掘和选别、提高资源综合利用效率均具有重要意义。

关键词: 黄铜矿, 深度学习, 矿物定量, 黄屯铜金矿床

Abstract:

Over the years, quantitative indicators in the upstream (exploration) to downstream (mining, beneficiation, and smelting) phases of the industrial chain have primarily relied on valuable elements. However, due to the diversity of minerals hosting these elements (where a single element can be present in multiple mineral phases), relying solely on elemental quantification often leads to significant discrepancies between resource reserve assessments in the early stages and resource recovery evaluations later on. Evidently, as the beneficiation process aims to recover target minerals containing recoverable elements, it is more reasonable to use the content of target minerals, rather than target elements, as the quantitative indicator. Consequently, for recoverable target minerals, the application of artificial intelligence image recognition technology for rapid and accurate quantification merits further investigation. The Huangtun copper-gold deposit in Anhui, a hydrothermal deposit, has copper (Cu) as one of its primary recoverable resources, with chalcopyrite being the primary mineral phase hosting Cu. Through the use of deep learning methods for microscopic image analysis, chalcopyrite was identified and quantified in 114 samples from 13 boreholes across five exploration lines in the mining area. The results demonstrate that AI image recognition technology based on deep learning can accurately identify and quantify chalcopyrite, providing a more comprehensive understanding of the spatial variation and distribution of copper resources compared to elemental quantification. This has significant implications for guiding exploration, mining, beneficiation, and separation of deep copper deposits or related minerals (such as gold), ultimately enhancing the comprehensive utilization efficiency of resources.

Key words: chalcopyrite, deep learning, mineral quantification, Huangtun copper-gold deposit

中图分类号:

P57
TP391.4

唐赧钰, 申俊峰, 陈强, 卿敏, 赵永建, 刘海明, 董博, 陈满, 龚智诚, 张萌萌, 林浩, 王浩阳. 基于深度学习的安徽黄屯铜金矿床黄铜矿定量分析及其意义[J]. 现代地质, 2025, 39(03): 541-551.

TANG Nanyu, SHEN Junfeng, CHEN Qiang, QING Min, ZHAO Yongjian, LIU Haiming, DONG Bo, CHEN Man, GONG Zhicheng, ZHANG Mengmeng, LIN Hao, WANG Haoyang. Quantitative Analysis of Chalcopyrite in Huangtun Copper-Gold Deposit in Anhui Province Based on Deep Learning and Its Significance[J]. Geoscience, 2025, 39(03): 541-551.

图/表 11

图1 黄屯铜金矿床区域地质图(底图据文献[33,36,28-30])

Fig.1 Regional geological map of Huangtun copper-gold deposit (after refs.[33,36,28-30])

图2 主要矿石、围岩和蚀变特征 (a)隐爆角砾岩筒内钻孔TK01样品;(b)隐爆角砾岩外钻孔ZK1802样品;(c)矿区南东远离已知角砾岩筒钻孔ZK1801样品;(d)下层粉砂岩;(e)上层闪长玢岩;(f)围岩电气石化和碳酸盐化;Ccp. 黄铜矿; Py. 黄铁矿; Qtz. 石英; Hb. 普通角闪石; Pl. 斜长石; Cal. 方解石; Tur. 电气石

Fig.2 Major ore, wall rock, and their alteration characteristics

图3 黄铜矿定量模型的代码结构简图

Fig.3 Schematic diagram of the code structure of a quantitative model of chalcopyrite

图4 偏光反光显微镜下不同赋存状态黄铜矿 (a)粗颗粒黄铜矿(Ccp);(b)细颗粒黄铜矿;(c)乳滴状黄铜矿被黄铁矿(Py)包围;(d)黄铜矿包含黄铁矿;(e)黄铜矿和黄铁矿连生;(f)粉末样中破碎的黄铜矿

Fig.4 Chalcopyrite in different occurrence states under polarized reflective microscopy

图5 黄屯主要勘探线位置及采样位置平面投影图(底图据文献[27])

Fig.5 Projection map of the main exploration line location and sampling location of Huangtun(after ref.[27])

图6 黄屯铜金矿区第18号勘探线采样位置图

Fig.6 Map of sampling location of Exploration Line 18 in Huangtun Copper-Gold Field

图7 DeeplabV3+整体构架 DCNN. 深度卷积神经网络层; ASPP. 带有空洞卷积的空间池化模块

Fig.7 Overall architecture of DeeplabV3+

图8 不同特征黄铜矿识别效果 (a)(c)(e)(g)不同特征黄铜矿(Ccp)及黄铁矿(Py)显微照片;(b)(d)(f)(h) 对应的图像识别结果

Fig.8 Identification effects of chalcopyrite with different characteristics

图9 TK01号钻孔岩芯14号样品部分视域黄铜矿定量结果

Fig.9 Quantitative results of chalcopyrite in partial field of view from sample of drill core 14 of TK01

图10 黄铜矿定量结果与Cu元素含量对比图

Fig.10 Comparison of chalcopyrite quantitative results and Cu

图11 18号线部分样品黄铜矿含量与Au元素关系

Fig.11 Relationship between chalcopyrite content and Au in some samples of Line 18

参考文献 39

[1]	OKADA N, MAEKAWA Y, OWADA N, et al. Automated identification of mineral types and grain size using hyperspectral imaging and deep learning for mineral processing[J]. Minerals, 2020, 10(9): 809.
[2]	张利军, 鲁文豪, 张建东, 等. 基于深度学习的镜下岩石、矿物薄片识别[J]. 地学前缘, 2024, 31(3): 498-510. DOI
[3]	BARKER R D, BARKER S L L, WILSON S, et al. Quantitative mineral mapping of drill core surfaces I: A method for μXRF mineral calculation and mapping of hydrothermally altered, fine-grained sedimentary rocks from a carlin-type gold deposit[J]. Economic Geology, 2021, 116(4): 803-819.
[4]	LIU H M, HARRIS J, SHERLOCK R, et al. Mineral prospectivity mapping using machine learning techniques for gold exploration in the Larder Lake area, Ontario, Canada[J]. Journal of Geochemical Exploration, 2023, 253: 107279.
[5]	VEZHAPPARAMBU V S, EIDSVIK J, ELLEFMO S L. Rock classification using multivariate analysis of measurement while drilling data: Towards a better sampling strategy[J]. Minerals, 2018, 8(9): 384.
[6]	WANG H C, FU T F, DU Y Q, et al. Scientific discovery in the age of artificial intelligence[J]. Nature, 2023, 620: 47-60.
[7]	ZUO R G, CARRANZA E J M. Machine learning-based mapping for mineral exploration[J]. Mathematical Geosciences, 2023, 55(7): 891-895.
[8]	武国朋. 基于机器学习的集宁浅覆盖区钼多金属矿成矿预测与评价[D]. 北京: 中国地质大学(北京), 2020.
[9]	张振杰, 成秋明, 杨玠, 等. 机器学习与成矿预测: 以闽西南铁多金属矿预测为例[J]. 地学前缘, 2021, 28(3): 221-235. DOI
[10]	周光锋. 基于机器学习的个旧地区锡铜多金属矿成矿预测[D]. 北京: 中国地质大学(北京), 2021.
[11]	郑超杰, 罗先熔, 刘攀峰. 应用机器学习在阿舍勒矿区外围的找矿预测[C]// 中国地球物理学会. 首届全国矿产勘查大会论文集, 2021:687.
[12]	刘颖. 基于机器学习的加拿大阿萨巴斯卡盆地东南部铀成矿预测与评价[D]. 抚州: 东华理工大学, 2022.
[13]	欧阳渊, 刘洪, 李光明, 等. 基于随机森林算法的找矿预测-以冈底斯成矿带西段斑岩-浅成低温热液型铜多金属矿为例[J]. 中国地质, 2023, 50(2): 303-330.
[14]	马瑶, 赵江南. 机器学习方法在矿产资源定量预测应用研究进展[J]. 地质科技通报, 2021, 40(1): 132-141.
[15]	郭海伟. 基于增量学习的储气库井底压力快速计算方法[J]. 断块油气田, 2025, 32(2): 292-299.
[16]	李辉, 阎建国, 陈榆桂, 等. CNN算法的损失函数优化及在低信噪比资料中的应用[J]. 断块油气田, 2023, 30(1): 107-113.
[17]	温利刚, 贾木欣, 王清, 等. 自动矿物学:Ⅱ.图像处理与自动测量[J]. 中国矿业, 2020, 29 (增2): 361-367.
[18]	陈倩, 宋文磊, 杨金昆, 等. 矿物自动定量分析系统的基本原理及其在岩矿研究中的应用: 以捷克泰思肯公司TIMA为例[J]. 矿床地质, 2021, 40(2): 345-368.
[19]	郝慧珍, 顾庆, 胡修棉. 基于机器学习的矿物智能识别方法研究进展与展望[J]. 地球科学, 2021, 46(9): 3091-3106.
[20]	谢小敏, 李利, 袁秋云, 等. 应用TIMA分析技术研究Alum页岩有机质和黄铁矿粒度分布及沉积环境特征[J]. 岩矿测试, 2021, 40(1): 50-60.
[21]	李秋杭, 谢远云, 康春国, 等. 基于人工和TIMA自动化方法的松花江水系重矿物组成: 对源-汇物源示踪的指示[J]. 海洋地质与第四纪地质, 2022, 42(3): 170-183.
[22]	OMOTOSO O, MCCARTY D K, HILLIER S, et al. Some successful approaches to quantitative mineral analysis as revealed by the 3rd Reynolds Cup contest[J]. Clays and Clay Minerals, 2006, 54(6): 748-760.
[23]	郭艳军, 周哲, 林贺洵, 等. 基于深度学习的智能矿物识别方法研究[J]. 地学前缘, 2020, 27(5): 39-47. DOI
[24]	周永章, 王俊, 左仁广, 等. 地质领域机器学习、深度学习及实现语言[J]. 岩石学报, 2018, 34(11): 3173-3178.
[25]	徐述腾, 周永章. 基于深度学习的镜下矿石矿物的智能识别实验研究[J]. 岩石学报, 2018, 34(11): 3244-3252.
[26]	李雷, 卢才武, 江松, 等. 基于改进ConvNeXt网络的矿物图像智能识别[J]. 地质通报, 2024, 43(7): 1266-1275.
[27]	王昭, 许强平, 张国强, 等. 安徽庐江黄屯金铜矿体空间形态特征及找矿方向[J]. 安徽地质, 2022, 32(3): 193-196, 208.
[28]	侯照方, 王安迪, 王昭, 等. 安徽庐江黄屯铜金矿床原生晕特征及矿床成因[J]. 华东地质, 2021, 42(3): 339-349.
[29]	谢杰, 张冠华, 吴文龙, 等. 安徽省庐江县黄屯铜金矿床地质特征及成因探讨[J]. 安徽地质, 2019, 29(4): 254-256, 264.
[30]	张矿, 赵进飞. 安徽省庐江县黄屯金铜矿控矿构造分析[J]. 世界有色金属, 2019(11): 133, 135.
[31]	许强平, 付光明, 周乾, 等. 庐枞盆地东北部马鞭山—黄屯—鹤毛河地区地球物理特征及找矿预测[J]. 华东地质, 2020, 41(2): 151-158.
[32]	胡生平, 韩善楚, 张洪求, 等. 庐枞盆地西湾铅锌矿床黄铁矿微量元素组成特征及成矿启示[J]. 现代地质, 2024, 38(1): 183-197.
[33]	毛景文, 段超, 刘佳林, 等. 陆相火山-侵入岩有关的铁多金属矿成矿作用及矿床模型: 以长江中下游为例[J]. 岩石学报, 2012, 28(1): 1-14.
[34]	周涛发, 范裕, 袁峰, 等. 庐枞盆地侵入岩的时空格架及其对成矿的制约[J]. 岩石学报, 2010, 26(9): 2694-2714.
[35]	周涛发, 范裕, 王世伟, 等. 长江中下游成矿带成矿规律和成矿模式[J]. 岩石学报, 2017, 33(11): 3353-3372.
[36]	王彪, 范裕, 周涛发, 等. 安徽庐枞盆地黄屯隐爆角砾岩筒型金铜矿床碲元素的分布、分配特征及沉淀机制研究[J]. 岩石学报, 2023, 39(10): 3048-3068.
[37]	ZHAO Y S, LIU H M, QIN K Z. Discussion on Au transportation mechanism in melt-magma-fluid in porphyry Cu deposit-A case study from Jinchang porphyry Au (Cu) deposit, Heilongjiang Province, China[J]. Ore Geology Reviews, 2019, 111: 102968.
[38]	宋伟全, 尤佳顺. 金矿床中共生组合及主要金属矿物特征[J]. 黑龙江科技信息, 2013(11): 26.
[39]	PENG H W, FAN H R, LIU X, et al. New insights into the control of visible gold fineness and deposition: A case study of the Sanshandao gold deposit, Jiaodong, China[J]. American Mineralogist, 2021, 106(1): 135-149.

基于深度学习的安徽黄屯铜金矿床黄铜矿定量分析及其意义

Quantitative Analysis of Chalcopyrite in Huangtun Copper-Gold Deposit in Anhui Province Based on Deep Learning and Its Significance

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 39

相关文章 3

编辑推荐

Metrics

本文评价

[1]	王永昌, 刘彩云, 熊杰, 王康, 胡焕发, 康佳帅. 基于生成对抗网络的半监督地震波阻抗反演[J]. 现代地质, 2024, 38(06): 1585-1593.
[2]	薛瑞洁, 熊杰, 张月, 王蓉. 基于卷积神经网络的磁异常反演[J]. 现代地质, 2023, 37(01): 173-183.
[3]	王庆, 曾齐红, 张友焱, 邵燕林, 魏薇, 邓帆. 基于多尺度区域卷积神经网络的露头孔洞自动提取[J]. 现代地质, 2021, 35(04): 1147-1154.