[1]彭 朋,高浪超,李家春.基于改进Yolov8-GCB的公路落石检测方法[J].长安大学学报(自然科学版),2025,45(2):24-35.[doi:10.19721/j.cnki.1671-8879.2025.02.003]
　PENG Peng,GAO Lang-chao,LI Jia-chun.Road rockfall detection method based on enhanced Yolov8-GCB[J].Journal of Chang’an University (Natural Science Edition),2025,45(2):24-35.[doi:10.19721/j.cnki.1671-8879.2025.02.003]
点击复制

基于改进Yolov8-GCB的公路落石检测方法()
分享到：

长安大学学报（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
第45卷
期数:
2025年2期
页码:
24-35
栏目:
道路工程
出版日期:
2025-03-31

文章信息/Info

Title:
Road rockfall detection method based on enhanced Yolov8-GCB
文章编号:
1671-8879(2025)02-0024-12
作者:
彭 朋1; 高浪超1; 李家春2
(1. 陕西科技大学 电气与控制工程学院,陕西 西安 710021; 2. 长安大学 公路学院,陕西 西安 710064)
Author(s):
PENG Peng1;  GAO Lang-chao1;  LI Jia-chun2
(1. School of Electrical and Control Engineering, Shannxi University of Science & Technology, Xi'an 710021,Shaanxi, China; 2.School of Highway, Chang'an University, Xi'an 710064, Shaanxi, China)
关键词:
道路工程;  落石检测;  Yolov8;  轻量化模型;  注意力机制;  目标检测
Keywords:
road engineering;  rockfall detection;  Yolov8;  lightweight model;  attention mechanism;  object detection
分类号:
U495
DOI:
10.19721/j.cnki.1671-8879.2025.02.003
文献标志码:
A
摘要:
为确保道路交通安全、及时检测并预警滚落到公路上的落石,构建面向落石灾害的数据集,设计一套基于机器视觉的落石检测系统,并提出一种改进的YOLOv8算法(YOLOv8-GCB),该算法在提升检测精度的同时,优化模型计算效率,便于部署于边缘计算设备。首先,在YOLOv8-GCB算法中设计幻影主干网络,将主干网络中的普通卷积单元替换为幻影卷积单元,降低模型的参数量和计算复杂度,提升模型在低算力设备上的运行效率。其次,在主干网络中引入通道空间混合注意力机制,使模型能够更好地关注落石的关键特征,增强对复杂背景的适应能力。最后,设计跨越加权融合网络,将跨越连接和加权融合的思想融入到特征融合网络中,进一步提升模型对不同尺度落石目标的检测性能,并与YOLOv8算法进行对比试验。研究结果表明:YOLOv8-GCB 算法的平均精度在AP@0.5上提高了1.2%,在AP@0.75上提升了1.1%,模型参数量下降了14.1%,模型计算量下降了16.1%; 上述改进为边缘设备在有限算力条件下实现公路落石灾害的智能检测提供了可行的技术解决方案,在确保检测精度的同时,有效兼顾了实时性与资源受限的双重约束,为公路落石灾害的智能化监测与预警奠定了技术基础。
Abstract:
To ensure road traffic safety and enable the timely detection and early warning of rockslides on highways, a dataset dedicated to rockfall hazards was constructed, and a machine vision-based rockfall detection system was developed. An improved YOLOv8 algorithm(YOLOv8-GCB)was proposed, which enhances detection accuracy while optimizing computational efficiency for deployment on edge computing devices. Firstly, a ghost convolution bottleneck network was designed in the YOLOv8-GCB algorithm, where standard convolutional units in the backbone network were replaced with ghost convolutional units, reducing the model's parameter size and computational complexity, and improving operational efficiency on low-resource devices. Secondly, a channel-spatial hybrid attention mechanism was introduced into the backbone network to enhance the model's focus on critical features of rockslides and strengthen its adaptability to complex backgrounds. Finally, a cross-layer weighted fusion network was designed by integrating cross-layer connections and weighted fusion into the feature fusion network, further improving detection performance for multi-scale rockfall targets. Comparative experiments with the original YOLOv8 algorithm demonstrated that YOLOv8-GCB achieved improvements of 1.2% in AP@0.5 and 1.1% in AP@0.75, while reducing the parameter size by 14.1% and computational load by 16.1%. The results indicate that the proposed method provides a viable technical solution for intelligent rockfall detection under limited computational resources on edge devices. It effectively balances real-time performance and resource constraints while maintaining detection accuracy, laying a technical foundation for intelligent monitoring and early warning systems for highway rockfall hazards.7 tabs, 12 figs, 29 refs.

参考文献/References:

[1] 王学良,刘海洋,王瑞琪,等.输变电工程崩塌(滚石)灾害识别与预测方法[J].工程地质学报,2018,26(1):172-178.
WANG Xue-liang,LIU Hai-yang,WANG Rui-qi,et al.The approach of rock collapse(rockfall)identification and prediction for power transmission and transformation project in mountain area[J].Journal of Engineering Geology,2018,26(1):172-178.
[2]王 栋,王剑锋,李天斌,等.西南山区某铁路隧道口高位落石三维运动特征分析[J].地质力学学报,2021,27(1):96-104.
WANG Dong,WANG Jian-feng,LI Tian-bin,et al.Analysis of three-dimensional movement characteristics of rockfall:A case study at a railway tunnel entrance in the southwestern mountainous area[J].Journal of Geomechanics,2021,27(1):96-104.
[3]胡厚田.崩塌落石研究[J].铁道工程学报,2005,22(增1):387-391.
HU Hou-tian.Research on rockfall[J].Journal of Railway Engineering,2005,22(S1):387-391.
[4]王 杰,叶 茂,马凤山等.基于视频图像识别的崩塌落石监测预警系统设计与实现[J].应用基础与工程科学学报,2014,22(5):952-963.
WANG Jie,YE Mao,MA Feng-shan,et al.Design and implementation of rockfall monitoring and early warning system based on video image recognition[J].Journal of Applied Basic and Engineering Sciences,2014,22(5):952-963.
[5]简云瑞,肖 硕.基于数字图像处理的边坡落 石识别算法研究[J].科技创新导报,2017,14(8):88-89.
JIAN Yun-rui,XIAO Shuo.Research on rockfall recognition algorithm based on digital image processing[J].Technology Innovation and Development Report,2017,14(8):88-89.
[6]刘林芽,吴送英,左志远,等.基于YOLOv3算法的山区铁路边坡落石检测方法研究[J].计算机科学,2021,48(增2):290-294.
LIU Lin-ya,WU Song-ying,ZUO Zhi-yuan,et al.Research on mountainous railway sidewall rockfall detection based on YOLOv3 algorithm[J].Computer Science,2021,48(S2):290-294.
[7]顾清华,杜艺凡,李萍丰,等.基于加权双向特征融合的矿区道路落石检测[J].黄金科学技术,2023,31(6):953-963.
GU Qing-hua,DU Yi-fan,LI Ping-feng,et al.Rockfall detection on roadways in mining areas based on weighted bidirectional feature fusion[J].Journal of Gold Science and Technology,2023 31(6):953-963.
[8]陈 垦,欧 鸥,杨长志,等.基于改进YOLOX的落石检测方法[J].计算机测量与控制,2023,31(11):13-59.
CHEN Ken,OU Ou,YANG Chang-zhi,et al.Rockfall detection method based on improved YOLOX[J].Computer Measurement and Control,2023,31(11):13-59.
[9]REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:Unified,real-time object detection[C]//IEEE.Proceedings of the IEEE conference on computer vision and pattern recognition.New York:IEEE,2016:779-788.
[10]LIU W,ANGUELOV D,ERHAN D,et al.Ssd:Single shot multibox detector[C]//ECCV.European Association for Computer Vision.Amsterdam:ECCV,2016:21-37.
[11]ROLET P,SEBAG M,TEYTAUD O.Integrated recognition,localization and detection using convolutional networks[C]//ECML.European Conference on Machine Learning.Bristol:ECML,2012:1255-1263.
[12]ALE L,ZHANG N,LI L.Road damage detection using RetinaNet[C]//IEEE.2018 IEEE International Conference on Big Data.New York:IEEE,2018:5197-5200.
[13]GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//IEEE.Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.New York:IEEE,2014:580-587.
[14]GIRSHICK R.Fast r-cnn[C]//IEEE.Proceedings of the IEEE International Conference on Computer Vision.New York:IEEE,2015:1440-1448.
[15]REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,39(6):1137-1149.
[16]HE K,GKIOXARI G,DOLLAR P,et al.Mask r-cnn[C]//IEEE.Proceedings of the IEEE international conference on Computer Vision.New York:IEEE,2017:2961-2969.
[17]XIAO B,NGUYEN M,YAN W Q.Fruit ripeness identification using YOLOv8 model[J].Multimedia Tools and Applications,2023,83(9):1-18.
[18]WANG C Y,BOCHKOVSKIY A,LIAO H Y M.YOLOv7:Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//IEEE.Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2023:7464-7475.
[19]BIST R B,SUBEDI S,YANG X,et al.A novel YOLOv6 object detector for monitoring piling behavior of cage-free laying hens[J].AgriEngineering,2023,5(2):905-923.
[20]WU T H,WANG T W,LIU Y Q.Real-time vehicle and distance detection based on improved YOLOV5 network[C]//IEEE.2021 3rd World Symposium on Artificial Intelligence(WSAI).New York:IEEE,2021:24-28.
[21]LIU W,ANGUELOV D,ERHAN D,et al.Ssd:Single shot multibox detector[C]//ECCV.Computer Vision-ECCV 2016:14th European Conference.Amsterdam:ECCV,2016:21-37.
[22]TAN M,PANG R,LE Q V.Efficientdet:Scalable and efficient object detection[C]//IEEE.Proceedings of the IEEE/CVF conference on computer vision and pattern recognition.New York:IEEE,2020:10781-10790.
[23]WANG C Y,LIAO H Y M,YEH I H et al.CSPNet:A new backbone that can enhance learning capability of CNN[C]//IEEE.2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops(CVPRW).New York:IEEE,2020:1571-1580.
[24]HE K M,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36:1904-1916.
[25]LIU S,LU Q,QIN H,et al.Path Aggregation Network for Instance Segmentation[C]//IEEE.2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2018:8759-8768.
[26]HAN K,WANG Y,TIAN Q,et al.Ghostnet:More features from cheap operations[C]//IEEE.Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2020:1580-1589.
[27]HUANG H,CHEN Z G,ZOU Y,et al.Channel prior convolutional attention for medical image segmentation[J].Computers in biology and medicine,2023,178:1080-1091.
[28]WOO S,PARK J,LEE J Y,et al.Cbam:Convolutional block attention module[C]//ECCV.Proceedings of the European Conference on Computer Vision.Amsterdam:ECCV,2018:3-19.
[29]TAN M,PANG R,LE Q V,et al.Efficientdet:Scalable and efficient object detection[C]//IEEE.Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2020:10781-10790.

相似文献/References:

[1]武建民,祝伟,马士让,等.应用加权密切值法评价基质沥青抗老化性能[J].长安大学学报(自然科学版),2012,32(01):0.
[2]张宜洛,袁中山.SMA混合料结构参数的影响因素[J].长安大学学报(自然科学版),2012,32(01):0.
[3]陈璟,袁万杰,郝培文,等.微观指标对沥青热稳定性能的影响[J].长安大学学报(自然科学版),2012,32(01):0.
[4]周兴业,刘小滔,王旭东,等.基于轴载谱的沥青路面累计当量轴次换算[J].长安大学学报(自然科学版),2012,32(01):0.
[5]李祖仲,王伯禹,陈拴发,等.轴载对复合式路面应力吸收层荷载应力的影响[J].长安大学学报(自然科学版),2012,32(01):0.
[6]关博文,刘开平,陈拴发,等.水镁石纤维路面混凝土路用性能[J].长安大学学报(自然科学版),2012,32(01):0.
[7]翁效林,王玮,张留俊,等.拓宽路基荷载下管桩复合地基沉降变形模式[J].长安大学学报(自然科学版),2012,32(01):0.
[8]穆柯,王选仓,柳志军,等.基于非饱和渗流原理的路基含水率预估[J].长安大学学报(自然科学版),2012,32(01):0.
[9]李振霞,陈渊召.不同类型半刚性基层材料性能的试验与分析[J].长安大学学报(自然科学版),2012,32(01):0.
[10]马 骉,马 晋,周宇鹏.沥青混合料降温收缩断裂特性[J].长安大学学报(自然科学版),2012,32(03):1.
　MA Biao,MA Jin,ZHOU Yu-peng.Cooling shrinkage fracture characteristic of asphalt mixture[J].Journal of Chang’an University (Natural Science Edition),2012,32(2):1.

备注/Memo

备注/Memo:
收稿日期:2024-09-12
基金项目:“十四五”重点研发计划项目(2022YFD2100600); 陕西省自然科学基础研究计划项目(2024JC-YBQN-0400); 陕西省教育厅项目(22JK0302)
作者简介:彭 朋(1990-),男,陕西安康人,讲师,硕士研究生导师,E-mail:pengpeng@sust.edu.cn。
更新日期/Last Update: 2025-04-01