Gold Science and Technology
img

Wechat

Adv. Search
Online Submission Peer Review Office Work

Gold Science and Technology ›› 2022, Vol. 30 ›› Issue (4): 603-611.doi: 10.11872/j.issn.1005-2518.2022.04.013

• Mining Technology and Mine Management • Previous Articles    

Research on Infrared Obstacle Detection of Mine Roadway Based on RCR_YOLOv4

Shunling RUAN1,2(),Lijuan DONG1(),Caiwu LU1,2,Qinghua GU1,2   

  1. 1.School of Resource Engineering, Xi’an University of Architecture and Technology, Xi’an 710055, Shaanxi, China
    2.Xi’an Key Laboratory of Intelligent Industry Perception Computing and Decision Making, Xi’an 710055, Shaanxi, China
  • Received:2022-01-05 Revised:2022-05-09 Online:2022-08-31 Published:2022-10-31
  • Contact: Lijuan DONG E-mail:ruanshunling@163.com;Dianalijuan1124@163.com

RichHTML

0

PDF (PC)

226

Abstract:

Aiming at the problem that driving obstacles such as falling rocks or pedestrians may appear on the dim road of underground mine roadway，which seriously affects the safe driving of unmanned mine card，a fast detection optimization model RCR_YOLOv4 of roadway obstacles based on infrared vision recognition was proposed.Firstly，the infrared camera was used for data acquisition，random cutting，random flipping，mirror flipping and other methods were used to expand the experimental data set.Labeling software was used for data Labeling，and the infrared obstacle data set of mine roadway was built and loaded into the obstacle detection model.Secondly，K-Means ++ optimization algorithm was used to screen the prior frame size of obstacles in the roadway，and depth separable convolution was introduced to reduce the number of network parameters and computation，so as to improve the positioning accuracy and detection efficiency of obstacle targets.The dual-channel attention mechanism is designed to optimize the network feature fusion module to realize the high-precision detection of the obstacle of unmanned mine jamming.The results show that the detection accuracy of the model can reach 93.52% and the detection speed can reach 60.6 FPS.Compared with the current popular target detection networks such as Faster_RCNN，SSD and YOLOv4，RCR_YOLOv4 also shows better comprehensive performance and can provide guarantee for the safe driving of unmanned mine cards in the complex environment of mine roadway.

Key words: mine roadway, driverless, machine vision, obstacle detection, infrared image, YOLOv4

CLC Number: 

  • TD421

Fig.1

RCR-YOLOv4 network structure of roadway obstacle detection model"

Fig.2

Main modules of the obstacle detection network model"

Table 1

Obstacle target detection effect after YOLOv4 clustering algorithm"

方法mAP/%检测速度/FPS
原始YOLOv490.1453.1
改进聚类算法的YOLOv491.3253.1

Fig.3

Standard convolution and depth separable convolution"

Fig.4

Residual module"

Fig.5

Convolution attention mechanism module"

Fig.6

Comparison of infrared image of mine roadway before and after pretreatment"

Table 2

Training parameters of obstacle detection model"

训练参数数值
迭代次数/次100
每次迭代的图片数量/张4
学习率0.001
训练样本数/个3 046

Fig.7

Loss curve of RCR_YOLOv4 model training"

Table 3

Comparison of different optimization performance of obstacle detection models"

模型深度可分离卷积RCRK-Means++

mAP

/%

检测速度

/FPS

模型大小

/M

Original_YOLOv490.1453.1243.9
优化模型1√88.2762.4114.9
优化模型2√93.6755.1272.5
优化模型3√91.3253.1243.9
优化模型4√√90.3460.4144.6
优化模型5√√89.7661.9114.9
优化模型6√√93.8655.1272.5
RCR_YOLOv4√√√93.5260.6144.6

Fig.8

Obstacle detection effect of different networks"

Fig.9

Comparison of losses of different models"

Table 4

Performance comparison of different detection network models"

模型

person AP

/%

stone AP

/%

mAP

/%

检测速度

/FPS

模型大小

/M

Faster_RCNN91.4048.3369.8747.6108.0
SSD98.7477.5588.1459.791.2
Original_YOLOv491.6788.6090.1453.1243.9
RCR_YOLOv493.6893.3693.5260.6144.6
Arthur D， Vassilvitskii S，2006.K-means++：The Advantages of Careful Seeding［M］. Palo Alto：Stanford InfoLab Publication.
Bi Lin， Xie Wei， Cui Jun，2017.Identification research on the miner’s safety helmet wear based on convolutional neural network［J］.Gold Science and Technology，25（4）：73-80.
Bochkovskiy A， Wang C Y， Liao H Y M，2020.Yolov4：Optimal speed and accuracy of object detection［J］.arXiv，2004.10934.
Cao M C， Wang J M，2020.Obstacle detection for autonomous driving vehicles with multi-lidar sensor fusion［J］.Journal of Dynamic Systems Measurement and Control，142（2）：021007.DOI：10.1115/1.4045361 .
doi: 10.1115/1.4045361
Chen Yi， Zhang Shuai， Wang Guiping，2020.Vehicle detection algorithm based on information fusion of LIDAR and camera［J］.Machinery & Electronics，38（1）：52-56.
Cui Tiejun， Wang Lingxiao，2021.Research on application of YOLOv4 object detection algorithm in monitoring on masks wearing of coal miners［J］.Journal of Safety and Technology，17（10）：66-71.
He K， Gkioxari G， Dollár P，et al，2017.Mask r-cnn［C］//IEEE International Conference on Computer Vision（ICCV） .Venice：IEEE： 2961-2969.
He K， Zhang X， Ren S，et al，2016.Deep residual learning for image recognition［C］//IEEE Conference on Computer Vision and Pattern Recognition（CVPR）.Las Vegas：IEEE： 770-778.
Howard A G， Zhu M， Chen B，et al，2017.Mobilenets：Efficient convolutional neural networks for mobile vision applications［J］.arXiv，1704.04861.
Li T P， Xu W H， Wang W，et al，2020.Obstacle detection in a field environment based on a convolutional neural network security［J］.Enterprise Information Systems，16（3）：472-493.
Liu W， Anguelov D， Erhan D，et al，2016.Ssd：Single shot multibox detector［C］//European Conference on Computer Vision（ECCV）.Amsterdam：IEEE： 21-37.
Lu Caiwu， Qi Fan， Ruan Shunling，2020.An open-pit mine roadway obstacle warning method intergrating the object detection and distance threshold model［J］.Opto-Electronic Engineering，47（1）：40-47.
Niu Lixia， Li Xiaomeng，2021.A new safety management model of intelligent mines in 5G era［J］.China Safety Science Journal，31（6）：29-36.
Ren S Q， He K M， Girshick R，et al，2017.Faster r-cnn：Towards real-time object detection with region proposal networks［J］.IEEE Transactions on Pattern Analysis & Machine Intelligence，39（6）：1137-1149.
Ruan Shunling， Li Shaobo， Lu Caiwu，et al，2021.Road negative obstacle detection in open-pit mines based on multi scale feature fusion［J］.Journal of China Coal Society，46（Supp.2）：1-10.
Wang Jinghua， Wang Liguan， Bi Lin，2021.Obstacle detection technology of mine electric locomotive driverless based on computer vision technology［J］.Gold Science and Technology，29（1）：136-146.
Woo S， Park J， Lee J Y，et al，2018.Cbam：Convolutional block attention module［C］//European Conference on Computer Vision（ECCV）.Munich：IEEE： 3-19.
Zhang X Y， Zhou M， Qiu P，et al，2019.Radar and vision fusion for the real-time obstacle detection and identification［J］.Industrial Robot，46（3）：391-395.
Zhao Liang， Hu Jie， Llu Han，et al，2021.Semantic segmentation based deep learning algorithm for 3D object detection from point clouds［J］.Cinese Journal of Lasers，48（17）：1-22.
Zhong Y， Wang J， Peng J，et al，2020.Anchor box optimization for object detection［C］//IEEE/CVF Winter Conference on Applications of Computer Vision（WACV）.Snowmass：IEEE： 1286-1294.
毕林，谢伟，崔君，2017.基于卷积神经网络的矿工安全帽佩戴识别研究［J］.黄金科学技术，25（4）：73-80.
陈毅，张帅，汪贵平，2020.基于激光雷达和摄像头信息融合的车辆检测算法［J］.机械与电子，38（1）：52-56.
崔铁军，王凌霄，2021.YOLOv4目标检测算法在煤矿工人口罩佩戴监测工作中的应用研究［J］.中国安全生产科学技术，17（10）：66-71.
卢才武，齐凡，阮顺领，2020.融合目标检测与距离阈值模型的露天矿行车障碍预警［J］.光电工程，47（1）：40-47.
牛莉霞，李肖萌，2021.5G时代智慧矿山安全管理新模式［J］.中国安全科学学报，31（6）：29-36.
阮顺领，李少博，卢才武，等，2021.多尺度特征融合的露天矿区道路负障碍检测［J］.煤炭学报，46（增2）：1-10.
王京华，王李管，毕林，2021.基于计算机视觉技术的矿井电机车无人驾驶障碍物检测技术［J］.黄金科学技术，29（1）：136-146.
赵亮，胡杰，刘汉，等，2021.基于语义分割的深度学习激光点云3D目标检测算法［J］.中国激光，48（17）：1-22.
[1] Jinghua WANG,Liguan WANG,Lin Bi. Obstacle Detection Technology of Mine Electric Locomotive Driverless Based on Computer Vision Technology [J]. Gold Science and Technology, 2021, 29(1): 136-146.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!

©2018 Gold Science and Technology　

Telephone：0931-8277791　

E-mail: hjkx@lzb.ac.cn Postcode：730000　

Powered by Beijing Magtech Co. Ltd

陇ICP备17001318号-1