基于函数改进的YOLOv3车辆检测与识别算法

doi:10.11959/j.issn.2096-6652.202301

Abstract

Abstract:

YOLOv3 algorithm is characterized by high detection accuracy and high speed in the aspect of target detection and recognition. It is outstanding among similar algorithms, but there are still obvious problems. The loss operation of the wide and high part of the loss function of this algorithm is not obvious to the distinction between the big target and the small target, which leads to the problem that the loss calculation is not accurate enough and the prediction box size of the small target is not accurate. In view of this defect, the loss function of YOLOv3 algorithm was improved, the wide-height coordinate error was modified into the form of proportion, the non-maximum suppression method of the original model was improved and the overlap threshold was changed from the fixed value to the form of attenuation function. Finally, the model was applied to vehicle detection. The experimental results showed that the improved model improved the accuracy rate, recall rate and F1 value to different degrees while basically not affecting the vehicle detection speed, making the overall performance of the model better than the original YOLOv3 model.

Key words: YOLOv3, loss function, non-maximum suppression, vehicle detection, vehicle recognition

CLC Number:

TP14

Huajie SONG,Lei ZHOU. Vehicle detection and recognition algorithm based on function improvement of YOLOv3[J]. Chinese Journal of Intelligent Science and Technology, 2023, 5(4): 535-542.

Figures/Tables 10

References 17

1	刘凯, 李浥东, 林伟鹏. 车辆再识别技术综述[J]. 智能科学与技术学报, 2020, 2(1): 10-25.
	LIU K, LI Y D, LIN W P. A survey on vehicle re-identification[J]. Chinese Journal of Intelligent Science and Technology, 2020, 2(1): 10-25.
2	SEVILLA-LARA L, SUN D Q, JAMPANI V, et al. Optical flow with semantic segmentation and localized layers[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2016: 3889-3898.
3	于晓明, 李思颖, 史胜楠. 混合高斯融合三帧差的运动目标检测改进算法[J]. 红外技术, 2019, 41(3): 256-261.
	YU X M, LI S Y, SHI S N. An improved algorithm for moving target detection using a Gaussian mixture with three-frame difference[J]. Infrared Technology, 2019, 41(3): 256-261.
4	SONG Z Y, ALI S, BOUGUILA N. Background subtraction using infinite asymmetric Gaussian mixture models with simultaneous feature selection[J]. IET Image Processing, 2020, 14(11): 2321-2332.
5	LUO J H, LIN C H. Pure FPGA implementation of an HOG based real-time pedestrian detection system[J]. Sensors(Basel), 2018, 18(4):1174.
6	曾娟, 李守义, 张洪昌. 图像分区域多特征融合斜向车辆检测算法研究[J]. 公路交通科技, 2020, 37(8): 99-107.
	ZENG J, LI S Y, ZHANG H C. Study on sub-region multi-feature image fusion oblique vehicle detection algorithm[J]. Journal of Highway and Transportation Research and Development, 2020, 37(8): 99-107.
7	GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of 2014 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2014: 580-587.
8	GIRSHICK R. Fast R-CNN[C]//Proceedings of 2015 IEEE International Conference on Computer Vision. Piscataway: IEEE Press, 2016: 1440-1448.
9	REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[C]//Proceedings of IEEE Transactions on Pattern Analysis and Machine Intelligence. Piscataway: IEEE Press, 2016: 1137-1149.
10	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.
11	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 779-788.
12	田思佳, 顾强, 胡蓉, 等. 一种基于深度学习的机械臂分拣方法[J]. 智能科学与技术学报, 2020, 2(3): 268-274.
	TIAN S J, GU Q, HU R, et al. A robot sorting method based on deep learning[J]. Chinese Journal of Intelligent Science and Technology, 2020, 2(3): 268-274.
13	刘肯, 何姣姣, 张永平, 等. 改进YOLO的车辆检测算法[J]. 现代电子技术, 2019, 42(13): 47-50.
	LIU K, HE J J, ZHANG Y P, et al. Improved YOLO vehicle detection algorithm[J]. Modern Electronics Technique, 2019, 42(13): 47-50.
14	REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2017: 6517-6525.
15	REDMON J, FARHADI A. YOLOv3: an incremental improvement[J].arXiv preprint, 2018, arXiv: 1804.02767.
16	阮激扬. 基于YOLO的目标检测算法设计与实现[D]. 北京: 北京邮电大学, 2019.
	RUAN J Y. Design and implementation of object detection algorithm based on YOLO[D]. Beijing: Beijing University of Posts and Telecommunications, 2019.
17	杨眷玉. 基于卷积神经网络的物体识别研究与实现[D]. 成都: 电子科技大学, 2016.
	YANG J Y. Research and implementation of object detection based on convolutional neural networks[D]. Chengdu: University of Electronic Science and Technology of China, 2016.

Metrics

Recommended 0

No Suggested Reading articles found!

类别	原模型				改进模型
类别	R	P	F1	mAP	R	P	F1	mAP
All	0.81	0.762	0.786	0.934	0.91	0.867	0.888	0.955
Sedan	0.808	0.78	0.794	0.95	0.931	0.87	0.899	0.96
SUV	0.81	0.743	0.775	0.903	0.91	0.856	0.882	0.919
Taxi	0.803	0.776	0.789	0.95	0.899	0.89	0.894	0.95
Bus	0.829	0.763	0.795	0.966	0.917	0.907	0.912	0.985
Van	0.8	0.69	0.741	0.901	0.893	0.812	0.851	0.961

算法	R	P	F1	mAP	检测速度/fps
YOLOv3	0.81	0.762	0.785	0.934	17
YOLOv3-spp	0.878	0.812	0.843	0.951	16.8
YOLOv3-tiny	0.77	0.75	0.76	0.898	18.3
Faster-RCNN	0.879	0.84	0.859	0.953	2.4
改进模型	0.91	0.867	0.888	0.955	16.7

类别	原模型				改进模型
类别	R	P	F1	mAP	R	P	F1	mAP
All	0.8	0.751	0.775	0.899	0.887	0.802	0842	0.935
Sedan	0.811	0.762	0.786	0.915	0.91	0.83	0.868	0.953
SUV	0.799	0.757	0.777	0.921	0.902	0.811	0.854	0.94
Taxi	0.81	0.748	0.778	0.901	0.88	0.801	0.839	0.942
Bus	0.812	0.76	0.785	0.89	0.887	0.79	0.836	0.92
Van	0.768	0.728	0.747	0.868	0.856	0.778	0.815	0.92

Vehicle detection and recognition algorithm based on function improvement of YOLOv3

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 10

References 17

Related Articles 0

Metrics

Recommended 0