Vehicle detection algorithm based on improved YOLOv3

被引：0

作者：

Chen W.-Y. ^{[1
,2
,3
,4
]}

Zhao H.-C. ^{[1
,2
,3
]}

Liu P.-F. ^{[1
,2
,3
]}

Fang J. ^{[1
,2
,3
]}

Sun H. ^{[1
,2
,3
,4
]}

机构：

[1] Key Laboratory of Opto-Electronic Information Processing, Chinese Academy of Sciences, Shenyang

[2] Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang

[3] Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences, Shenyang

[4] University of Chinese Academy of Sciences, Beijing

来源：

Kongzhi yu Juece/Control and Decision | 2024年 / 39卷 / 04期

关键词：

coordinate attention; deep learning; KIoU Loss; soft-SPP; vehicle detection; YOLOv3;

D O I：

10.13195/j.kzyjc.2022.1044

中图分类号：

TB18 [人体工程学]; Q98 [人类学];

学科分类号：

030303 ; 1201 ;

摘要：

Aiming at the problems of vehicle detection in the traffic scene such as a large number of small targets and severe target occlusion, a single-stage target detection algorithm based on the improved YOLOv3 is proposed. Since the small target only contains fewer pixels and features are not obvious, this algorithm builds a Soft-SPP structure based on the idea of spatial pyramid pooling, which integrates multiple receptive fields and adopts soft-pooling operation to retain details to the maximum extent and avoids information loss. The coordinate attention mechanism is introduced to capture the remote dependence with accurate location information. and adjust the weight assigned to each channel feature to make the network better learn important information. A loss function KIoU Loss based on key points and aspect ratio is proposed as the boundary box loss function, which makes the boundary box regression more accurate. The experimental results show that the mAP of the improved algorithm on the autopilot KITTI data set is 94.69 %, which is 4.13 % higher than that of the original YOLOv3 algorithm, and the detection speed is only reduced by 3.16 frame•s−1, which significantly improves the detection accuracy while maintaining the detection speed. © 2024 Northeast University. All rights reserved.

引用

页码：1151 / 1159

页数：8

共 24 条

[1]

Li H, Fu K, Yan M L, Et al., Vehicle detection in remote sensing images using denoizing-based convolutional neural networks, Remote Sensing Letters, 8, 3, pp. 262-270, (2017)

[2]

Ren S Q, He K M, Girshick R, Et al., Faster R-CNN: Towards real-time object detection with region proposal networks, IEEE Transactions on Pattern Analysis and Machine Intelligence, 39, 6, pp. 1137-1149, (2017)

[3]

Luo J Q, Fang H S, Shao F M, Et al., Multi-scale traffic vehicle detection based on faster R-CNN with NAS optimization and feature enrichment, Defence Technology, 17, 4, pp. 1542-1554, (2021)

[4]

Nguyen H., Improving faster R-CNN framework for fast vehicle detection, Mathematical Problems in Engineering, 2019, pp. 1-11, (2019)

[5]

Zhou T, Li Z, Zhang C., Enhance the recognition ability to Occlusions and small objects with robust faster R-CNN, International Journal of Machine Learning and Cybernetics, 10, 11, pp. 3155-3166, (2019)

[6]

Kong F F, Song B B., Improved YOLOv3 panoramic traffic monitoring target detection, Computer Engineering and Applications, 56, 8, pp. 20-25, (2020)

[7]

Wang Y S, Jia W J, Wang C F, Et al., Vehicle recognition method based on improved YOLOv3 algorithm, Laser & Optoelectronics Progress, 58, 16, pp. 240-247, (2021)

[8]

Song S Q, Li X, Zhu X F, Et al., Urban Road vehicle detection method by aerial photography based on improved SSD, Transducer and Microsystem Technologies, 40, 1, pp. 114-117, (2021)

[9]

Redmon J, Farhadi A., YOLOv3: An incremental improvement, (2018)

[10]

Hou Q B, Zhou D Q, Feng J S., Coordinate attention for efficient mobile network design, IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 13708-13717, (2021)

← 1 2 3 →