基于深度学习轻量化的改进SSD煤矸快速分选模型

doi:10.12068/j.issn.1005-3026.2023.10.014

东北大学学报（自然科学版） ›› 2023, Vol. 44 ›› Issue (10): 1474-1480.DOI: 10.12068/j.issn.1005-3026.2023.10.014

基于深度学习轻量化的改进SSD煤矸快速分选模型

李娟莉^1,2，魏代良^1,2，李博^1,2，文小^1,2

(1. 太原理工大学机械与运载工程学院，山西太原030024; 2. 太原理工大学煤矿综采装备山西省重点实验室，山西太原030024)

发布日期:2023-10-27
通讯作者: 李娟莉
作者简介:李娟莉(1979-)，女，山西寿阳人，太原理工大学教授.
基金资助:
国家自然科学基金资助项目(51875386，51804207).

Improved SSD Rapid Separation Model of Coal Gangue Based on Deep Learning and Light-Weighting

LI Juan-li^1,2， WEI Dai-liang^1,2， LI Bo^1,2， WEN Xiao^1,2

1. College of Mechanical and Vehicle Engineering， Taiyuan University of Technology， Taiyuan 030024， China; 2. Shanxi Key Laboratory of Fully-Mechanized Coal Mining Equipment， Taiyuan University of Technology， Taiyuan 030024， China.

Published:2023-10-27
Contact: LI Bo
About author:-
Supported by:
-

摘要/Abstract

摘要： 针对SSD目标检测模型参数量大、运行速率低的问题，在SSD模型的基础上提出一种新的煤矸快速识别模型DSR-SSD.应用深度可分离卷积代替主干特征提取网络中的普通卷积，减少了模型的计算量；将RFB模块融入到SSD模型中，提高了模型的特征提取能力.经验证，DSR-SSD模型的识别速率为113.99帧/s、精确率为95.17%.将DSR-SSD与SSD，Faster-RCNN，YOLOv3三种模型对比，发现DSR-SSD模型与SSD模型相比，精确率提高了2.29%，识别速率提高了60.89%；同时，DSR-SSD模型的精确率比Faster-RCNN模型高2.86%，比YOLOv3模型高2.71%，识别速率分别是Faster-RCNN模型和YOLOv3模型的14.90倍和3.65倍，证明了DSR-SSD模型性能优越.

关键词: 煤矸分选；深度学习；目标检测；SSD模型；轻量化

Abstract: A new model DSR-SSD for coal gangue fast identification is proposed based on the SSD model to address the issues of large parameter quantities and low operating speed in the SSD model. The application of deep separable convolutions in the backbone feature extraction network reduces the computational complexity， and integrating the RFB module into the SSD model improves the model’s feature extraction ability. After verification， the recognition rate of the DSR-SSD model is 113.99 frames/s， and the accuracy rate is 95.17%. Comparing DSR-SSD with SSD， Faster-RCNN， and YOLOv3 models， it is found that the DSR-SSD model improves the accuracy by 2.29% and the recognition rate by 60.89% compared to the SSD model， and the accuracy of the DSR-SSD model is 2.86% higher than the Faster-RCNN and 2.71% higher than the YOLOv3， with recognition rates 14.90 and 3.65 times higher than the Faster-RCNN and YOLOv3.

Key words: coal gangue separation; deep learning; target detection; SSD model; light-weighting

中图分类号:

TD94

李娟莉，魏代良，李博，文小. 基于深度学习轻量化的改进SSD煤矸快速分选模型[J]. 东北大学学报（自然科学版）, 2023, 44(10): 1474-1480.

LI Juan-li， WEI Dai-liang， LI Bo， WEN Xiao. Improved SSD Rapid Separation Model of Coal Gangue Based on Deep Learning and Light-Weighting[J]. Journal of Northeastern University(Natural Science), 2023, 44(10): 1474-1480.

参考文献

[1]张文静，马喜立.以环境税治理雾霾的减排效果及减排成本——基于动态多区域CGE模型［J］.北京理工大学学报(社会科学版)，2020，22(3):36-47.(Zhang Wen-jing，Ma Xi-li.Emission reduction effect and emission reduction cost of haze by environmental tax—a dynamic multi-region CGE analysis［J］.Journal of Beijing Institute of Technology (Social Sciences Edition)，2020，22(3):36-47.)
[2]曹亦俊，刘敏，邢耀文，等.煤矿井下选煤技术现状和展望［J］.采矿与安全工程学报，2020，37(1):192-201.(Cao Yi-jun，Liu Min，Xing Yao-wen，et al.Current situation and prospect of underground coal preparation technology［J］.Journal of Mining & Safety Engineering，2020，37(1):192-201.)
[3]宋亮，刘善军，毛亚纯，等.基于可见光-近红外光谱的煤种分类方法［J］.东北大学学报(自然科学版)，2017，38(10):1473-1476.(Song Liang，Liu Shan-jun，Mao Ya-chun，et al.Coal classification based on visible and near-infrared spectrum［J］.Journal of Northeastern University (Natural Science)，2017，38(10):1473-1476.)
[4]Wang W D，Lu Z Q，Lu H R.Research on methods to differentiate coal and gangue using image processing and a support vector machine［J］.International Journal of Coal Preparation and Utilization，2021，41(8):603-616.
[5]Dou D Y，Zhou D Y，Yang J，et al.Coal and gangue recognition under four operating conditions by using image analysis and relief-SVM［J］.International Journal of Coal Preparation and Utilization，2018，40(7):1-10.
[6]Wang X Q，Wang S，Guo Y，et al.Dielectric and geometric feature extraction and recognition method of coal and gangue based on VMD-SVM［J］.Powder Technology，2021，392(7):241-250.
[7]Xie Y T，Chi X W，Li H J，et al.Coal and gangue recognition method based on local texture classification network for robot picking［J］.Applied Sciences，2021，11(23):100-113.
[8]Pu Y Y，Apel D B，Szmigiel A，et al.Image recognition of coal and coal gangue using a convolutional neural network and transfer learning［J］.Energies，2019，12(9):1735-1745.
[9]Bai F Y，Fan M Q，Yang H P，et al.Fast recognition using convolutional neural network for the coal particle density range based on images captured under multiple light sources［J］.International Journal of Mining Science and Technology，2021，(31):1053-1061.
[10]Ren S Q，He K M，Girshick R，et al.Faster R-CNN:towards real-time object detection with region proposal networks［J］.IEEE Transactions on Pattern Analysis & Machine Intelligence，2017 39(6):1137-1149.
[11]Liu W，Anguelov D，Erhan D，et al.SSD:single shot multibox detector［C］//European Conference on Computer Vision.Tel Aviv，Israel:Springer，2016:21-37.
[12]Redmon J，Farhadi A.YOLOv3:an incremental improvement［J］.ArXiv，2018，7:18-24.
[13]单鹏飞，孙浩强，来兴平，等.基于改进Faster R-CNN的综放煤矸混合放出状态识别方法［J］.煤炭学报，2022，47(3):1382-1394.(Shan Peng-fei，Sun Hao-qiang，Lai Xing-ping，et al.Identification method on mixed and release state of coal-gangue masses of fully mechanized caving based on improved Faster R-CNN［J］.Journal of China Coal Society，2022，47(3):1382-1394.)
[14]Li D Y，Wang G F，Zhang Y，et al.Coal gangue detection and recognition algorithm based on deformable convolution YOLOv3［J］.IET Image Processing (Wiley-Blackwell)，2021，16(1):134-144.
[15]来文豪，周孟然，胡锋，等.基于多光谱成像和改进YOLOv4的煤矸石检测［J］.光学学报，2020，477(24):72-80.(Lai Wen-hao，Zhou Meng-ran，Hu Feng，et al.Coal gangue detection based on multi-spectral imaginig and improved YOLOv4［J］.Acta Optica Sinica，2020，477(24):72-80.)
[16]Liu Q，Li J G，Li Y S，et al.Recognition methods for coal and coal gangue based on deep learning［J］.IEEE Access，2021，PP(99):77599-77610.
[17]Shafiee M J，Chywl B，Li F，et al.Fast YOLO:a fast you only look once system for real-time embedded object detection in video［J］.Journal of Computational Vision and Imaging Systems，2017，3(1):18-20.
[18]宋欣，李奇，解婉君，等.YOLOv3-ADS:一种基于YOLOv3的深度学习目标检测压缩模型［J］.东北大学学报(自然科学版)，2021，42(5):609-615.(Song Xin，Li Qi，Xie Wan-jun，et al.YOLOv3-ADS:a compression model for deep learning object detection based on YOLOv3［J］.Journal of Northeastern University (Natural Science)，2021，42(5):609-615.)

基于深度学习轻量化的改进SSD煤矸快速分选模型

Improved SSD Rapid Separation Model of Coal Gangue Based on Deep Learning and Light-Weighting

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

编辑推荐

Metrics

本文评价