摘要
目的:交通标志的检测是车载辅助系统的关键环节之一,针对YOLO V3检测算法得到的检测结果存在目标框不精确的问题,提出改进交通标志目标检测算法。方法:YOLO V3是当前目标检测算法中检测召回率高且速度较优的算法,但在定位上不够准确。为解决该问题,本文采用流行排序算法对得到的检测框进行二次修正,从而使得目标定位精度提升。结果:通过结合YOLO V3和流行排序算法使目标检测框的交并比提升了3%~9%。结论:通过YOLO V3和Ranking Saliency的结合能够使得目标检测的定位精度提高。
Aims:The detection algorithm is the key part of a vehicle-assisted system.Aiming at the problem of inaccurate detection target frame obtained by YOLO V3 detection algorithm,an algorithm for improving traffic sign target detection was proposed.Methods:YOLO V3 has high recall rate and the fastest speed in the current target detection algorithm,but the positioning is not accurate.In order to solve this problem,this paper proposed to use the Ranking Saliency algorithm to correct the predicted position of the target.Results:By combining the YOLO V3 and Ranking Saliency algorithms,the IoU of the target prediction box increased by 3%-9%.Conclusions:The experiments show that the proposed algorithm can improve the target positioning accuracy.
引文
[1]LEVINSON J,ASKELAND J,BECKER J,et al.Towards fully autonomous driving:Systems and algorithms[C]//2011IEEE Intelligent Vehicles Symposium(Ⅳ).Baden-Baden:IEEE,2011:163-168.
[2]TIMOFTE R,PRISACARIU V A,GOOL L V,et al.Combining Traffic Sign Detection with 3DTracking Towards Better Driver Assistance[M].Hanckensack,USA:World Scientific Publishing,2011:425-446.
[3]KAMADA H,NAOI S,AND GOTOH T.A compact navigation system using image processing and fuzzy control[C]//IEEE Proceedings on Southeastcon.New Orleans:IEEE,1990:337-342.
[4]ESCALERA A,MORENO L,ARMINGOL J,et al.Road traffic sign detection and classification[J].IEEE Transactions on Industrial Electronics,1997,44(6):848-859.
[5]MIURA J,KANDA T,SHIRAI Y.An active vision system for real-time traffics sign recognition[C]//ITSC2000.2000IEEE Intelligent Transportation Systems.Dearborn:IEEE,2000:52-57.
[6]ARNOUL P,VIALA M,GUERIN J,et al.Traffic signs localization for highways inventory from a video camera on board a moving collection van[C]//Proceedings of Conference on Intelligent Vehicles.Tokyo:IEEE,1996:141-146.
[7]MALDONADO-BASCON S,LAFUENTE-ARROYO S,GIL-JIMENEZ P,et al.Road-sign detection and recognition based on support vector machines[J].IEEE Transaction on Intelligent Transportation System,2007,8(2):264-278.
[8]LIU H,LIU D,XIN J.Real-time recognition of road traffic sign in motion image based on genetic algorithm[C]//Proceedings of the 2021International Conference on Machine Learning and Cybernetics.Beijing:IEEE,2002:83-86.
[9]MATHIAS M,TIMOFTE R,BENENSON R,et al.Traffic sign recognition-how far are we from the solution[C]//Proceedings of the 2013IEEE International Joint Conference on Neural Networks.Dallas:IEEE,2013:1-8.
[10]HE K,ZHANG X,REN S,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,27(9):1904-1916.
[11]GIRSHICK R.FAST R-CNN[C]//Proceedings of the2015IEEE International Conference on Computer Vision.Piscataway:IEEE,2015:1440-1448.
[12]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&Machine Intelligence,2017,39(6):1137-1149.
[13]REDMON J,FARHADI A.YOLO9000:Better,faster,stronger[C]//IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Honolulu:IEEE,2017:6517-6525.
[14]LIU W,ANGELOV D,ERHAN D,et al.SSD:single shot multi box detector[C]//Proceedings of the European Conference on Computer Vision,LNCS 9905.Heidelberg:Springer Berlin,2016:21-37.
[15]JIN J Q,FU K,ZHANG C.Traffic sign recognition with hinge loss trained convolutional neural network[J].IEEETransactions on Intelligent Transportation Systems,2014,15(5):1991-2000.
[16]QIAN R,ZHANG B,YUE Y,et al.Robust Chinese traffic sign detection and recognition with deep convolutional neural network[C]//Proceedings of the 2015International Conference on Natural Computation.Piscataway:IEEE,2015:791-796.
[17]ZHU Z,LIANG D,ZHANG S,et al.Traffic-sign detection and classification in the wild[C]//The IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Las Vegas:IEEE,2016:2110-2118.
[18]CHUAN Y,LIHE Z,HUCHUAN L,et al.Saliency detection vis graph-based manifold ranking[C]//The IEEEConference on Computer Vision and Pattern Recognition(CVPR).Portland:IEEE,2013:3166-3173.
[19]LIN TY,DOLLAR P,GIRSHICK R,et al.Feature pyramid networks for object detection[C]//The IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Honolulu:IEEE,2017:2117-2125.