Traffic sign detection algorithm based on improved YOLOv7

Xiao Zhang; Zhenyu Zhang

doi:10.1117/12.2681271

8 June 2023 Traffic sign detection algorithm based on improved YOLOv7

Xiao Zhang, Zhenyu Zhang

Proceedings Volume 12707, International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2023); 1270751 (2023) https://doi.org/10.1117/12.2681271
Event: International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2023), 2023, Changsha, China

Abstract

Aiming at the problems of low detection accuracy and slow detection speed of existing deep learning-based traffic sign detection methods, this paper proposes an improved YOLOv7 traffic sign detection method CN-YOLOv7. First, based on the idea of ConvNeXt network, this paper introduces ConvNeXt_block with integrated attention in the feature extraction network to reduce the loss of small target traffic sign feature information during downsampling, improve the accuracy of traffic sign detection and reduce the number of network floating point operations; in the detection head part, in order to improve the detection ability for tiny targets, this paper adds a small target detection head. On this basis, in order to speed up the network calculation, this paper adjusts the maximum pooling layer arrangement of the SPPCSPC module based on the detection layer structure to obtain the SPPFCSPC module, which replaces the SPPCSPC module and reduces the number of network parameters under the same feeling field. In addition, this paper uses K-means++ clustering algorithm to obtain predefined anchor frames to make them more suitable for the dataset, which improves the robustness and detection accuracy of the algorithm. The comparison experiments on the homemade dataset and CCTSDB dataset show that compared with the YOLOv7 algorithm, the CN-YOLOv7 algorithm improves the mAP by 4.1% and the inference speed reaches 61.7 FPS. Compared with SSD, Faster R-CNN, YOLOv5, YOLOv7 and other methods, the CN-YOLOv7 ensures the detection speed At the same time, it can effectively improve the detection accuracy, and the detection effect is better than the original network model and the traditional classical target detection network model.

Citation Download Citation

Xiao Zhang and Zhenyu Zhang "Traffic sign detection algorithm based on improved YOLOv7", Proc. SPIE 12707, International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2023), 1270751 (8 June 2023); https://doi.org/10.1117/12.2681271

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
9 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Object detection

Target detection

Convolution

Feature extraction

Small targets

Deep learning

Head

Keywords/Phrases

Search In:

Publication Years