基于轻量化网络和迁移学习的岩石智能识别
DOI:
作者:
作者单位:

山西大学数学科学学院

作者简介:

通讯作者:

中图分类号:

P588

基金项目:

国家自然科学基金项目(编号:61976128;62072293);山西省基础研究计划资助项目(编号:202303021221054); 山西省回国留学人员科研教研资助项目(编号:2024-002);山西省研究生教育教学改革课题(编号:2022YJJG010)


Intelligent rock recognition based on lightweight network and transfer learning
Author:
Affiliation:

School of Mathematical Sciences

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    在岩石图像识别中,实现岩石快速准确的识别是岩石数字化发展的关键。其中,光照、湿度等环境因素引起的图像模糊问题成为岩石智能识别的最大挑战之一。基于此,本文提出了一种新的深度学习方法(MbileNetV3-small-RegNetX)来识别岩石图像,其适用于移动设备等资源有限的场景。本文在RegNet网络的基础上采用迁移学习方法,结合MobileNetV3残差结构与通道注意力(Squeeze-and-Excitation,SE)模块的优势,有效地优化了特征提取与网络结构,并显著提升了检测速度。为验证该方法的准确性,本文将新模型与当下主流的轻量化模型(DenseNet和ShuffleNet)进行消融对比实验。结果显示,本研究的新模型表现出高精度(82.15%)、快速(0.06 GFLOP)的特点。此外,该模型对于光照、湿度等环境因素引起的图像模糊具有良好的适应性。

    Abstract:

    In rock image recognition, achieving rapid and accurate identification of rocks is crucial for the digitalization of rocks. Among the challenges faced in intelligent rock recognition is the issue of image blurring caused by environmental factors such as lighting and humidity. In light of this, a novel deep learning approach (MobileNetV3-small-RegNetX) is proposed in this paper for rock image recognition, which is suitable for scenarios with limited resources such as mobile devices. Building upon the RegNet network, the study employs transfer learning methods, combining the advantages of the MobileNetV3 residual structure with SE (Squeeze-and-Excitation,SE)modules to effectively optimize feature extraction and network structure, leading to a significant improvement in detection speed. To validate the accuracy of this approach, comparative experiments are conducted between the new model and current mainstream lightweight models (DenseNet and ShuffleNet). The results demonstrate that the new model proposed in this study exhibits high precision (82.15%) and fast processing (0.06 GFLOP). Additionally, the model demonstrates good adaptability to environmental factors such as lighting and humidity-induced image blurring.

    参考文献
    相似文献
    引证文献
引用本文

李顺勇,李青辉,邢煜曼. 基于轻量化网络和迁移学习的岩石智能识别[J]. 科学技术与工程, , ():

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2024-01-30
  • 最后修改日期:2024-05-25
  • 录用日期:2024-06-05
  • 在线发布日期:
  • 出版日期:
×
亟待确认版面费归属稿件,敬请作者关注