基于改进YOLOv4的X射线图像违禁品检测算法

doi:10.3969/j.issn.1000-1093.2021.12.015

摘要/Abstract

摘要： 为提高安检速度、实现X射线图像中违禁物品的自动检测，提出一种基于改进YOLOv4的X射线图像违禁品检测算法。该算法在单阶段目标检测算法YOLOv4基础上设计一种空洞密集卷积模块。将上采样链路融合后特征输入空洞密集卷积模块中，增强特征表达能力和卷积视野。对融合后特征信息加入注意力机制，用来增强有效特征和抑制无效特征，最终得到表征图像信息的特征图输入检测头部。采用Mosaic数据增强方法训练网络，提升网络的鲁棒性。结果表明：该算法在公开SIXray数据集上的均值平均精度达到80.16%，检测速度为25帧/s；该算法在公开SIXray数据集上多类违禁物品能够取得较高的检测精度，且满足检测的实时性要求。

关键词: 违禁品检测, YOLOv4, X射线图像, 空洞密集卷积, 注意力机制, 数据增强

Abstract: An improved YOLOv4 algorithm for detecting the prohibited items in X-ray images is proposed to increase the speed of security inspection and realize the automatic detection of prohibited items in X-ray images. The proposed algorithm is used to design a dilated dense convolution module based on the one-stage object detection algorithm YOLOv4. The features after the upsampling link fusion are input into the dilated dense convolution module to enhance the feature expression ability and the convolution field of vision. An attention mechanism is added to the fused feature information to enhance effective features and suppress invalid features. Finally，a feature map representing image information is input to detection head. Mosaic data enhancement method is used to train the network to improve the robustness of the network. The results show that the mean average precision (mAP) of the proposed algorithm on the public SIXray data set reaches 80.16%，and the detection speed is 25 frames per second (FPS). The proposed algorithm can achieve high detection accuracy for multiple types of prohibited items on the public SIXray dataset， and meet the real-time requirements of detection.

Key words: prohibiteditemsdetection, YOLOv4, X-rayimage, dilateddenseconvolution, attentionmechanism, dataaugmentation

中图分类号:

TP391.4

穆思奇，林进健，汪海泉，魏雄志. 基于改进YOLOv4的X射线图像违禁品检测算法[J]. 兵工学报, 2021, 42(12): 2675-2683.

MU Siqi， LIN Jinjian， WANG Haiquan， WEI Xiongzhi. An Algorithm for Detection of Prohibited Items in X-ray Images Based on Improved YOLOv4[J]. Acta Armamentarii, 2021, 42(12): 2675-2683.

参考文献

［1］徐茂书.注意力机制下X光安检违禁品图像区域分割[D].天津：中国民航大学，2019.
XU M S.Prohibited item X-ray image segmentation via attention mechanism[D].Tianjin：Civil Aviation University of China，2019.(in Chinese)
[2］ DUMAGPI J K，JUNG W Y，JEONG Y J. A new GAN-based anomaly detection (GBAD) approach for multi-threat object classification on large-scale X-ray security images[J]. IEICE Transactions on Information and Systems，2020，E103.D(2):454-458.
[3］赵子豪.基于GAN的安检X光物品图像生成方法研究[D]. 天津：中国民航大学，2019.
ZHAO Z H. Research on security X-ray item image generation based on GAN[D]. Tianjin：Civil Aviation University of China，2019.(in Chinese)
[4］ LIN T Y，MAIRE M，BELONGIE S，et al. Microsoft COCO: common objects in context[C]∥Proceedings of European Conference on Computer Vision.Zurich，Switzerland: Springer，2014：740-755.
[5］ WU X，SAHOO D，HOI SC H.Recent advances in deep learning for object detection[J].Neurocomputing，2020，396：39-64.
[6］ JIAO L，ZHANG F，LIU F，et al. A survey of deep learning-based object detection[J].IEEE Access，2019，7:128837-128868.
[7］ WEI Y L，TAO R S，WU Z J，et al. Occluded prohibited items detection: an X-ray security inspection benchmark and de-occlusion attention module[C]∥Proceedings of the 28th ACM International Conference on Multimedia. Seattle，WA，US:ACM，2020:138-146.
[8］张友康，苏志刚，张海刚，等.X光安检图像多尺度违禁品检测[J].信号处理，2020，36(7):1096-1106.
ZHANG Y K，SU Z G，ZHANG H G，et al. Multi-scale prohibited item detection in X-ray security image[J].Journal of Signal Processing，2020，36(7): 1096-1106.(in Chinese)
[9］郭守向，张良.Yolo-C:基于单阶段网络的X光图像违禁品检测[J].激光与光电子学进展，2021，58（8）：0810003.
GUO S X，ZHANG L.Yolo-C: one-stage network for prohibited items detection within X-ray images[J]. Laser & Optoelectronics Progress, 2021，58（8）：0810003. (in Chinese)

[10］ BOCHKOVSKIY A，WANG C Y，LIAO H.YOLOv4: Optimal speed and accuracy of object detection[EB/OL]. (2020-04-23) [2020-09-16]. https:∥arxiv.org/abs/ 2004.10934.
[11］原培新，张晓慧.数字图像处理在汽车轮胎 X 射线检测中的应用[J].CT理论与应用研究，2007，16(2):48-51.
YUAN P X，ZHANG X H. X-ray detection of tyre with digital image processing[J]. Computerized Tomography Theory and Applications，2007，16(2):48-51.(in Chinese)
[12］顾乐旭.基于X射线安检图像的刀具识别系统设计[D].吉林：东北电力大学，2018.
GU L X. Design of cutlery recognition system based on X-ray safety inspection image[D].Jilin：Northeast Electric Power University，2018.(in Chinese)
[13］ NERCESSIAN S，PANETTA K，AGAIAN S. Automatic detection of potential threat objects in X-ray luggage scan images[C]∥Proceedings of IEEE Conference on Technologies for Homeland Security.Waltham，MA，US: IEEE，2008: 504-509.
[14］韩宁.基于深度学习的X射线图像危险品检测与跟踪算法研究[D].兰州：兰州大学，2018.
HAN N.A deep learning-based dangerous goods detection and tracking algorithm from X-ray images[D]. Lanzhou：Lanzhou University，2018.(in Chinese)
[15］ LIU L J，LENG X，LIU Y. Deep convolutional neural network based object detector for X-ray baggage security imagery[C]∥Proceedings of the 2019 IEEE 31st International Conference on Tools with Artificial Intelligence. Portland， OR，US:IEEE，2019：1757-1761.
[16］ AKCAY S，KUNDEGORSKI M E，WILLCOCKS C G，et al. Using deep convolutional neural network architectures for object classification and detection within X-ray baggage security imagery[J]. IEEE Transactions on Information Forensics and Security，2018，13(9):2203-2215.
[17］ WANG P Q，CHEN P F，YUAN Y，et al.Understanding convolution for semantic segmentation[C]∥Proceedings of the 2018 IEEE Winter Conference on Applications of Computer Vision. Lake Tahoe，NV，US: IEEE，2018：1451-1460.
[18］ WANG Q L，WU B G，ZHU P F，et al.ECA-Net: efficient channel attention for deep convolutional neural networks[C]∥Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle，WA，US: IEEE，2020:11531-11539.
[19］ LI X，WANG W H，HU X L，et al.Selective kernel networks[C]∥
Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach， CA，US:IEEE，2019:510- 519.
[20］ ZHANG Z，WANG P，LIU W，et al. Distance-IoU loss: faster and better learning for bounding box regression [EB/OL]. (2019-11-19) [2020-02-28]. https:∥arxiv.org/abs/1911.08287?context=cs.CV.
[21］ MIAO C J，XIE L X，WAN F，et al.SIXray: a large-scale security inspection X-ray benchmark for prohibited item discovery in overlapping images[C]∥Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Long Beach，CA，US:IEEE，2019：2114-2123.
[22］ LIU W， ANGUELOV D，ERHAN D，et al.SSD: single shot MultiBox detector[EB/OL].(2015-12-08) [2019-10-12]. https:∥arxiv.org/abs/1512.02325.
[23］ REDMON J，FARHADI A.YOLOv3: an incremental improvement [EB/OL].(2018-04-08)[2020-01-12]. https:∥arxiv.org/abs/1804.02767.

[1]	林森，王金刚，高宏伟. 基于全局补偿注意力机制的战场图像去雾方法[J]. 兵工学报, 2024, 45(4): 1344-1353.
[2]	熊光明, 罗震, 孙冬, 陶俊峰, 唐泽月, 吴超. 基于红外相机和毫米波雷达融合的烟雾遮挡无人驾驶车辆目标检测与跟踪[J]. 兵工学报, 2024, 45(3): 893-906.
[3]	宋晓茹, 刘康, 高嵩, 陈超波, 阎坤. 复杂战场环境下改进YOLOv5军事目标识别算法研究[J]. 兵工学报, 2024, 45(3): 934-947.
[4]	熊佳梅, 王永振, 燕雪峰, 魏明强. 一种基于语义引导和对比学习的战场图像去烟算法[J]. 兵工学报, 2024, 45(2): 671-683.
[5]	吕卫民, 孙晨峰, 任立坤, 赵杰, 李永强. 一种基于TCN-LGBM的航空发动机气路故障诊断方法[J]. 兵工学报, 2024, 45(1): 253-263.
[6]	秦昊林, 许廷发, 李佳男. 基于超像素注意力和孪生结构的半监督高光谱显著性目标检测[J]. 兵工学报, 2023, 44(9): 2639-2649.
[7]	卞紫阳, 许廷发, 马亮, 李佳男. 基于边缘智能终端的跨模态行人重识别方法及应用[J]. 兵工学报, 2023, 44(9): 2631-2638.
[8]	李曾琳, 李波, 白双霞, 孟波波. 基于AM-SAC的无人机自主空战决策[J]. 兵工学报, 2023, 44(9): 2849-2858.
[9]	彭沛然, 任术波, 李佳男, 周鸿伟, 许廷发. 基于光照感知的多光谱融合行人检测方法[J]. 兵工学报, 2023, 44(9): 2622-2630.
[10]	吴礼洋, 呙鹏程, 刘超, 李文强. 基于注意力机制增强残差网络的雷达信号调制类型识别[J]. 兵工学报, 2023, 44(8): 2310-2318.
[11]	连静, 丁荣琪, 李琳辉, 王雪成, 周雅夫. 基于图模型和注意力机制的车辆轨迹预测方法[J]. 兵工学报, 2023, 44(7): 2162-2170.
[12]	李萌, 武海军, 董恒, 任光, 张鹏, 黄风雷. 基于机器学习的混凝土侵彻深度预测模型[J]. 兵工学报, 2023, 44(12): 3771-3782.
[13]	陈子涵. 基于多模态Transformer的机电作动器剩余寿命预测[J]. 兵工学报, 2023, 44(10): 2920-2931.
[14]	杨春伟, 刘炳琪, 王继平, 邵节, 韩治国. 基于注意力机制的高超声速飞行器LSTM智能轨迹预测[J]. 兵工学报, 2022, 43(S2): 78-86.
[15]	张博尧，冷雁冰. 基于YOLOv4网络模型的金属表面划痕检测[J]. 兵工学报, 2022, 43(S1): 214-221.