Vehicle Logo Recognition with YOLOv5 Location and Multi-Feature Fusion

doi:10.3778/j.issn.1002-8331.2207-0389

Abstract

Abstract: In order to solve the problem of vehicle logo identification in intelligent transportation system, a vehicle logo image recognition scheme based on using YOLOv5s network to locate vehicle logo and multi-feature fusion is proposed. The YOLOv5s network is selected in the vehicle logo positioning stage to meet the requirements of the positioning speed and accuracy. Firstly, in the vehicle logo recognition stage, the vehicle logo edges with different effects are obtained by adjusting the parameters in the extended difference of Gaussians. Then, a set of two-dimensional Gabor filters are designed to filter the edge-detected vehicle logo image and extract the corresponding feature vector. Finally, by calculating the Euclidean distance between the feature vector of the vehicle logo image to be tested and the feature vector in the standard comparison library, the label index corresponding to the smallest distance is taken as the classification and recognition result. The optimal recognition accuracy of the scheme is 96.91%, which can reach 99.33% after classification with random forest algorithm. The best overall accuracy of vehicle logo positioning and recognition in this paper exceeds that by using the YOLOv5s network in one step, and it is significantly improved compared with the traditional image processing method.

Key words: vehicle logo identification, YOLOv5s, multi-feature fusion, extended difference of Gaussians, two dimensional Gabor filtering, Euclidean distance, random forest

摘要： 为解决智能交通系统中车标识别的问题，提出YOLOv5s网络车标定位多特征融合的车标图像识别方案。车标定位阶段选择YOLOv5s网络以满足对车标定位速度与精度等的需求。车标识别阶段通过调整扩展高斯差分中的参数得到具有不同效果的车标边缘，设计一组二维Gabor滤波器对边缘检测后的车标图像进行滤波处理并提取出对应的车标图像特征向量，通过计算待测车标图像特征与标准比对库中特征向量的欧几里德距离，取距离最小者对应的标签索引作为分类识别结果，该方案的最佳识别正确率为96.91%。采用随机森林算法进行分类后的最佳识别正确率可达99.33%。该方案的车标定位与识别最佳整体正确率超过了YOLOv5s网络直接一步到位识别车标的方案，且相较于传统图像处理方法有明显提升。

关键词: 车标识别, YOLOv5s, 多特征融合, 扩展高斯差分, 二维Gabor滤波, 欧几里德距离, 随机森林

DONG Guanghui, CHEN Xingyu. Vehicle Logo Recognition with YOLOv5 Location and Multi-Feature Fusion[J]. Computer Engineering and Applications, 2023, 59(5): 176-193.

董光辉, 陈星宇. YOLOv5定位多特征融合的车标识别[J]. 计算机工程与应用, 2023, 59(5): 176-193.

References

[1] LIU Y，LI S.A vehicle-logo location approach based on edge detection and projection[C]//Proceedings of 2011 IEEE International Conference on Vehicular Electronics and Safety，2011：165-168.
[2] SUN Q，LU X，CHEN L，et al.An improved vehicle logo recognition method for road surveillance images[C]//2014 Seventh International Symposium on Computational Intelligence and Design，2014：373-376.
[3] 刘玉松，陈刚.基于边缘投影的快速车标定位方法[J].信息技术，2015（11）：102-104.
LIU Y S，CHEN G.A fast vehicle-logo location method based on projection of edge[J].Information Technology，2015（11）：102-104.
[4] 张栋冰.基于视觉注意机制与支持向量机结合的车标定位方法[J].南京大学学报（自然科学），2018，54（2）：481-487.
ZHANG D B.Vehicle logo detection based on visual attention theory and SVM[J].Journal of Nanjing University（Natural Science），2018，54（2）：481-487.
[5] 焦志全.基于非限制场景ALPR系统的车标定位[J].农业装备与车辆工程，2020，58（10）：72-75.
JIAO Z Q.Vehicle logo location based on unconstrained scenarios ALPR system[J].Agricultural Equipment & Vehicle Engineering，2020，58（10）：72-75.
[6] 王枚，王国宏，高小林，等.基于PCA和边缘不变矩的车标识别新方法[J].计算机工程与应用，2008，44（4）：224-226.
WANG M，WANG G H，GAO X L，et al.New method of vehicle-logo recognition based on PCA and edge invariant moment[J].Computer Engineering and Applications，2008，44（4）：224-226.
[7] DAI S，HUANG H，GAO Z，et al.Vehicle-logo recognition method based on Tchebichef moment invariants and SVM[C]//2009 WRI World Congress on Software Engineering，2009：18-21.
[8] SOON F C，KHAW H Y，CHUAH J H.Pattern recognition of vehicle logo using Tchebichef and Legendre moment[C]//2015 IEEE Student Conference on Research and Develop-ment（SCOReD），2015：82-86.
[9] GU Q，YANG J，CUI G，et al.Multi-scale vehicle logo recognition by directional dense SIFT flow parsing[C]//2016 IEEE International Conference on Image Processing（ICIP），2016：3827-3831.
[10] 李熙莹，吕硕，江倩殷，等.基于Edge Boxes的大型车辆车标检测与识别[J].计算机工程与应用，2018，54（12）：152-159.
LI X Y，LV S，JIANG Q Y，et al.Large vehicle logo detection and recognition based on Edge Boxes[J].Computer Engineering and Applications，2018，54（12）：152-159.
[11] 李哲，于梦茹.基于多种LBP特征集成学习的车标识别[J].计算机工程与应用，2019，55（20）：134-138.
LI Z，YU M R.Vehicle-logo recognition based on ensemble learning with multiple LBP features[J].Computer Engineering and Applications，2019，55（20）：134-138.
[12] 曲爱妍，吴秋玲，张正，等.基于车标区域的智能识别方法研究[J].软件工程，2021，24（10）：37-40.
QU A Y，WU Q L，ZHANG Z，et al.Research on intelligent recognition method based on vehicle logo area[J].Software Engineering，2021，24（10）：37-40.
[13] PUNARSELVAM E，SURESH P.Edge detection of CT scan spine disc image using Canny edge detection algorithm based on magnitude and edge length[C]//3rd International Conference on Trendz in Information Sciences & Computing（TISC2011），2011：136-140.
[14] ROE E，MELLO C A B.Binarization of color historical document images using local image equalization and XDoG[C]//2013 12th International Conference on Document Analysis and Recognition，2013：205-209.
[15] 刘丽霞，李宝文，王阳萍，等.改进Canny边缘检测的遥感影像分割[J].计算机工程与应用，2019，55（12）：54-58.
LIU L X，LI B W，WANG Y P，et al.Remote sensing image segmentation based on improved Canny edge detection[J].Computer Engineering and Applications，2019，55（12）：54-58.
[16] DU B，HAO Z，WEI X.Roundness detection of end face for shaft workpiece based on Canny-Zernike sub pixel edge detection and improved Hough transform[C]//2021 IEEE 11th International Conference on Electronics Information and Emergency Communication（ICEIEC），2021：1-4.
[17] 齐小祥，李敏，朱颖，等.基于边缘检测的SAR图像自适应区域分割[J].计算机工程与应用，2021，57（22）：232-240.
QI X X，LI M，ZHU Y，et al.Adaptive region segmentation of SAR image based on edge detection[J].Computer Engineering and Applications，2021，57（22）：232-240.
[18] LV Wentao，YU Qiuze，YU Wenxian.Water extraction in SAR images using GLCM and support vector machine[C]//IEEE 10th International Conference on Signal Processing Proceedings，2010：740-743.
[19] MIHRETEAB K，IWAHASHI M，YAMAMOTO M.Crow birds detection using HOG and CS-LBP[C]//2012 International Symposium on Intelligent Signal Processing and Communications Systems，2012：406-409.
[20] FENG K，YUAN F.Static hand gesture recognition based on HOG characters and support vector machines[C]//2013 2nd International Symposium on Instrumentation and Measurement，Sensor Network and Automation（IMSNA），2013：936-938.
[21] MIRZAPOUR F，GHASSEMIAN H.Using GLCM and gabor filters for classification of PAN images[C]//2013 21st Iranian Conference on Electrical Engineering（ICEE），2013：1-6.
[22] VERMA K，KHUNTETA A.Facial expression recognition using Gabor filter and multi-layer artificial neural network[C]//2017 International Conference on Information，Communication，Instrumentation and Control（ICICIC），2017：1-5.
[23] 刘文培，李凤莲，张雪英，等.面向人脸识别的WPD-HOG金字塔特征提取方法[J].计算机工程与应用，2018，54（22）：150-155.
LIU W P，LI F L，ZHANG X Y，et al.WPD-HOG pyra-mid feature extraction method for face recognition[J].Computer Engineering and Applications，2018，54（22）：150-155.
[24] CHANG J，FU Z，LI N，et al.A novel ship detection method based on gabor feature[C]//IET International Radar Conference（IET IRC 2020），2020：858-862.
[25] 包文歧，谢立强，徐才华，等.基于YOLOv5的微小型无人机实时探测方法[J].兵器装备工程学报，2022，43（5）：232-237.
BAO W Q，XIE L Q，XU C H，et al.Real-time detection method of micro UAV based on YOLOv5[J].Journal of Ordnance Equipment Engineering，2022，43（5）：232-237.
[26] 邱天衡，王玲，王鹏，等.基于改进YOLOv5的目标检测算法研究[J].计算机工程与应用，2022，58（13）：63-73.
QIU T H，WANG L，WANG P，et al.Research on object detection algorithm based on improved YOLOv5[J].Computer Engineering and Applications，2022，58（13）：63-73.
[27] 沈科，季亮，张袁浩，等.基于改进YOLOv5s模型的煤矸目标检测[J].工矿自动化，2021，47（11）：107-111.
SHEN K，JI L，ZHANG Y H，et al.Research on coal and gangue detection algorithm based on improved YOLOv5s model[J].Industry and Mine Automation，2021，47（11）：107-111.
[28] 孙凯明，郝明，刘彤军，等.基于高斯差分滤波的金刚线颗粒检测方法[J].自动化技术与应用，2021，40（7）：121-124.
SUN K M，HAO M，LIU T J，et al.Diamond wire particle detection method based on DoG filtering[J].Techniques of Automation and Applications，2021，40（7）：121-124.
[29] WINNEM?LLER H，KYPRIANIDIS J E，OLSEN S C.XDoG：an extended difference-of-Gaussians compendium including advanced image stylization[J].Computers & Graphics，2012：740-753.
[30] DAUGMAN J G.Uncertainty relation for resolution in space，spatial frequency，and orientation optimized by two-dimensional visual cortical filters[J].Journal of the Optical Society of America A，1985，2（7）：1160-1169.
[31] 孔锐，张冰.Gabor滤波器参数设置[J].控制与决策，2012，27（8）：1277-1280.
KONG R，ZHANG B.Design of Gabor filters’ parameter[J].Control and Decision，2012，27（8）：1277-1280.
[32] KAMARAINEN J，KYRKI V，KALVIAINEN H.Fundamental frequency Gabor filters for object recognition[C]//2002 International Conference on Pattern Recognition，2002：628-631.
[33] YANG S，BO C，ZHANG J，et al.VLD-45：a big dataset for vehicle logo recognition and detection[J].IEEE Transactions on Intelligent Transportation Systems，2022，23（12）：25567-25573.