高徐辉 1黄宋魏 1陈永春 1钟婷婷 1何济帆 1吴丽萍 1程贯瑞1
作者信息
- 1. 昆明理工大学 国土资源工程学院,云南 昆明 650032
- 折叠
摘要
Abstract
With the promotion of smart mining construction in China,mineral resource development is gradually moving towards mechanization,automation,informatization,and intelligence.In recent years,the application of deep learning technology in mineral recognition has received widespread attention,especially image recognition technology based on deep learning,which provides new approach for mineral detection and classification.This article briefly described the main technologies and methods of deep learning in mineral recognition,introduced the current application status of deep learning technology in mineral classification,segmentation,particle size analysis,and summarized its application progress in mineral processing,and pointed out the problems of deep learning in mineral recognition.At present,the optimization direction of deep learning models applied to mineral image recognition mainly focuses on fea-ture extraction optimization,model architecture optimization,training strategy optimization,loss function optimiza-tion,interpretability,and visualization tools.Model optimization can improve the recognition ability of the model for the morphology,texture,and boundaries of mineral particles,accelerate the convergence speed of the model,enhance the generalization ability of the model,and improve the transparency and credibility of the model.关键词
深度学习/卷积神经网络/矿物识别/矿物分类/矿物分割/矿物粒度/数据集/模型优化Key words
deep learning/convolutional neural network/mineral identification/mineral classification/mineral seg-mentation/mineral particle size/data set/model optimization分类
天文与地球科学
