[1]张 震,陈可鑫,陈云飞.优化聚类和引入 CBAM 的 YOLOv5 管制刀具检测[J].郑州大学学报(工学版),2023,44(05):40-45.[doi:10.13705/j.issn.1671-6833.2022.05.015]
　ZHANG Zhen,CHEN Kexin,CHEN Yunfei.YOLOv5 with Optimized Clustering and CBAM for Controlled Knife Detection[J].Journal of Zhengzhou University (Engineering Science),2023,44(05):40-45.[doi:10.13705/j.issn.1671-6833.2022.05.015]
点击复制

优化聚类和引入 CBAM 的 YOLOv5 管制刀具检测()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
44
期数:
2023年05期
页码:
40-45
栏目:
出版日期:
2023-08-20

文章信息/Info

Title:
YOLOv5 with Optimized Clustering and CBAM for Controlled Knife Detection
作者:
张 震1;  陈可鑫2;  陈云飞1
1. 郑州大学 电气工程学院,河南 郑州 450001;2. 郑州大学 计算机与人工智能学院, 河南 郑州 450001
Author(s):
ZHANG Zhen1;  CHEN Kexin2;  CHEN Yunfei1
 1. School of Information Engineering,Chang′an University, Xi′an 710018, China; 2. National Computer Network Emergency Response Technical Team / Coordination Center of China, Beijing 100029, China
关键词:
管制刀具检测;  公共安全;  目标检测;  聚类算法;  注意力机制
Keywords:
controlled knife detection;  public safety;  target detection;  clustering algorithm;  attention mechanism
分类号:
:TP391
DOI:
10.13705/j.issn.1671-6833.2022.05.015
文献标志码:
A
摘要:
针对传统管制刀具检测方式过度依赖特定设备与环境,目标检测存在适用范围小、抗干扰能力弱、精度 低等问题,提出一种优化聚类和引入卷积注意力模块(CBAM)的 YOLOv5 管制刀具检测方法。 首先,通过计算每 个点的局部密度和该点与其他具有较高密度点之间的最小距离,绘制决策图并获取聚类中心;其次,计算局部类 中所有点与其类中心间的平均距离来提取核心点,并使用全局搜索分配策略将分类点归类;最后,采用统计学习 策略分配剩余点,未被处理的点当作噪声点,归入到其最近的类中。 采用改进密度峰值聚类算法对管制刀具的边 界框聚类进行分析,优化先验框尺寸,提高先验框与目标物体尺寸的匹配度,解决 YOLOv5 模型中 K-means 聚类 算法聚类效果不稳定以及对大规模数据收敛较慢的问题。 此外,将 Backbone 中的 C3 模块与 CBAM 注意力机制 相结合,改进为 CBAMC3 模块,提升模型对目标特征的提取能力,解决 YOLOv5 算法对小目标检测效果不佳的问 题,提高模型精度。 实验结果表明:改进后模型 YOLOv5-Plus 在自定义数据集上的 P、R、mAP@ 0. 5、mAP@ 0. 5 ∶ 0. 95 等参数的数值分为 98. 14%、95. 80%、97. 56%、76. 68%,相比于改进前的 YOLOv5 模型分别提高了 1. 64、 1. 59、1. 51、3. 26 百分点,有效提升了公共区域的管制刀具检测能力。
Abstract:
Aiming at the problems that traditional controlled knife detection methods relied too much on specific equipment and environment, and the target detection was applicable to small scope of application, with weak anti-interference ability and low precision, a YOLOv5 model with optimized clustering and CBAM was proposed for controlled knife detection was proposed. Firstly, a decision diagram could be drew and cluster centers could be obtained by calculating the local density of each point and the minimum distance between it and others with higher density. Then the core points could be extracted by calculating all points in the local class the average distance with its class center, the global search assignment strategy was used to classify the test points. Finally, the statisticallearning strategy was employed to allocate the remaining points, these unprocessed points were used as noise points, and they were classified into the class of its nearest neighbor. The improved density peak clustering algorithm was used to analyze the bounding box of the controlled knife, optimize the size of the priori box, and improve the matching degree between the priori box and the size of target object, so as to solve the problem that the clustering effect of the K-means clustering algorithm in the YOLOv5 model such as its unstability and the low convergence rate of large-scale data. Moreover,The C3 module in the Backbone network was combined with the CBAM attention mechanism, named CBAMC3 module, which could improve the model′s ability to extract target features, solve the problem that the YOLOv5 algorithm was not effective for small target detection, and improve the model accuracy. The experimental results showed that the values of P, R, mAP @ 0. 5, mAP @ 0. 5 ∶ 0. 95 of the improved model YOLOv5-Plus on the customed data set were 98. 14%, 95. 80%, 97. 56%, and 76. 68%, respectived, which is 1. 64%, 1. 59%, 1. 51%, and 3. 26% higher than that of YOLOv5 before improvement, and also verified that the proposed model could effectively improve the detection performance of controlled knife in public areas.

参考文献/References:

[1] 徐晨玉. 基于 YOLOv3 的危险物品检测模型研究与改 进[D] . 兰州: 兰州大学, 2019. 

XU C Y. Research and improvement of a dangerous goods detection model based on YOLOv3 [ D ] . Lanzhou: Lanzhou University, 2019. 
[2] 杨希. 基于深度学习的地铁安检危险物品检测[ D] . 武汉: 武汉理工大学, 2020. 
YANG X. Detection of dangerous goods in subway security check based on deep learning[D] . Wuhan: Wuhan University of Technology, 2020.
 [3] 王胜. 毫米波图像中危险物品检测[ D] . 北京: 清华 大学, 2018. 
 WANG S. Dangerous objects detection in millimeter wave images[D]. Beijing: Tsinghua University, 2018.
 [4] 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 and Machine Intelligence, 2017, 39(6): 1137-1149.
 [5] HE K M, GKIOXARI G, DOLLÁR P, et al. Mask R-CNN [C]∥2017 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2017: 2980-2988. 
[6] HE K M, ZHANG X Y, REN S Q, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[ J] . IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9) : 1904-1916.
 [7] FERNANDEZ-CARROBLES M, DENIZ O, MAROTO F. Gun and knife detection based on faster R-CNN for video surveillance[C]∥Iberian Conference on Pattern Recognition and Image Analysis. Cham: Springer International Publishing,2019: 441-452.
 [8] BOCHKOVSKIY A, WANG C Y, LIAO H Y M. Yolov4: optimal speed and accuracy of object detection [EB / OL] . (2020-08-23) [2021-09-13] . https:∥ doi. org / 10. 48550 / arXiv. 2004. 10934. 
[9] Ultralytics. YOLOv5[ EB / OL] . ( 2020- 05- 18) [ 2021- 08-12] . https:∥ github. com / ultralytics/ yolov5. 
[10] LIU W, ANGUELOV D, ERHAN D, et al. Ssd: single shot multibox detector [ C ] ∥ European Conference on Computer Vision. Berlin:Springer Press, 2016: 21-37.
 [11] 张伟彬, 吴军, 易见兵. 基于 RFB 网络的特征融合管 制物品检测算法研究[ J] . 广西师范大学学报( 自然 科学版) , 2021, 39(4) : 34-46. ZHANG W B, WU J, YI J B. Research on feature fusion controlled items detection algorithm based on RFB network[ J] . Journal of Guangxi Normal University ( Natural Science Edition) , 2021, 39(4) : 34-46.
 [12] YUN S, HAN D, CHUN S, et al. CutMix: regularization strategy to train strong classifiers with localizable features [C]∥2019 IEEE / CVF International Conference on Computer Vision ( ICCV) . Piscataway: IEEE, 2019: 6022 -6031.
 [13] 张震, 李浩方, 李孟洲, 等. 改进 YOLOv3 算法与人 体信息数据融合的视频监控检测方法[ J] . 郑州大学 学报(工学版) , 2021, 42(1) : 28-34. 
ZHANG Z, LI H F, LI M Z, et al. Video surveillance detection method based on improved YOLOv3 algorithm and human body information data fusion [ J] . Journal of Zhengzhou University ( Engineering Science) , 2021, 42 (1) : 28-34.
[14] WANG C Y, MARK LIAO H Y, WU Y H, et al. CSPNet: a new backbone that can enhance learning capability of CNN[ C]∥2020 IEEE / CVF Conference on Computer Vision and Pattern Recognition Workshops ( CVPRW ) . Piscataway: IEEE, 2020: 1571-1580. 
[15] RODRIGUEZ A, LAIO A. Clustering by fast search and find of density peaks [ J] . Science, 2014, 344 ( 6191) : 1492-1496. 
[16] 薛小娜, 高淑萍, 彭弘铭, 等. 结合 k 近邻的改进密 度峰值聚类算法[ J] . 计算机工程与应用, 2018, 54 (7) : 36-43. 
XUE X N, GAO S P, PENG H M, et al. Improved density peaks clustering algorithm combining k-nearest neighbors[ J] . Computer Engineering and Applications, 2018, 54(7) : 36-43. 
[17] WOO S, PARK J, LEE J Y, et al. Cbam: convolutional block attention module[C]∥Proceedings of the European Conference on Computer Vision (ECCV) . Berlin:Springer Press,2018: 3-19.

更新日期/Last Update: 2023-09-04