Abstract
A traffic sign image segmentation algorithm based on improved spatio-temporal graph convolution is proposed by fusing octave convolution and spatio-temporal component graph convolution network for the problem of road traffic sign recognition in a complex environment. The algorithm uses octave convolution to reduce the computational effort to improve the recognition speed and uses a spatio-temporal graph convolution network to recognize traffic signs more accurately. First, the acquired images are processed by data image enhancement; then, the RGB image saliency detection module based on octave convolution is used; then, the spatio-temporal map convolution network module is improved by using the SETR algorithm to train a lightweight and high-precision spatio-temporal map convolution network model; finally, the image details are optimized by using the octave convolution residual module and eventually used for road sign recognition. The experimental results show that the algorithm can effectively improve the segmentation accuracy rate, which is 16.5%, 10.1%, 6.1%, and 5.1%, respectively, compared with other algorithms; in terms of recognition speed, the single image processing time of different data sets is better than other algorithms; its recognition effect is also better than other algorithms in terrible weather conditions, such as intense light, fog, heavy rain, night and snow conditions, especially in intense light, heavy rain and low contrast weather conditions. In the ablation comparison experiments, its algorithm improves 12.5%, 7.3%, and 8.6% in segmentation accuracy compared with other module combination algorithms in the same data set case.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Wang, K., Zhao, Y., Xing, X.: Deep learning in driverless vehicles. CAAI Trans. Intell. Syst. 13(1), 55–69 (2018)
Chen, F., Liu, Y., Li, S.: Survey of traffic sign detection and recognition methods in complex environment. Comput. Eng. Sci. 57(16), 65–73 (2021)
Deng, X., Zhang, Y., Yang, Y.: A shape recognition algorithm for traffic sign identification. Comput. Eng. Sci. 43(02), 322–328 (2021)
Xu, J., Wang, L.: Traffic sign recognition method based on AlexNet network. RadioEngineering 1–10 (2022). http://kns.cnki.net/kcms/detail/13.1097.TN.20220111.1637.006.html
He, R.-B., Di, L., Liang, J.: An improved deep learning algorithm for road traffic sign recognition. CAAI Trans. Intell. Syst. 15(6), 1121–1130 (2020)
Dewi, C., Chen, R.-C., Tai, S.-K.: Evaluation of robust spatial pyramid pooling based on convolutional neural network for traffic sign recognition system (2020)
Cao, J., Song, C., Peng, S., et al.: Improved traffic sign detection and recognition algorithm for intelligent vehicles (2019)
Yazdan, R., Varshosaz, M., Pirasteh, S., Remondino, F.: Using geometric constraints to improve performance of image classifiers for automatic segmentation of traffic signs. https://doi.org/10.1139/geomat-2020-0010
Di, L., He, R., Liang, J.: Road traffic identification based on probability clustering and convolutional neural network. J. Nanjing Univ. (Nat. Sci.) 55(02), 238–250 (2019)
Jie-Wei, Li-Wei-Sang, Li-Wei: Traffic signs real-time classification and recognition of traffic signs based on multi-feature fusion. Mod. Electron. Technol. 42(11), 50–53+58 (2019)
Mannan, A., Javed, K., Rehman, A.U., et al.: Optimized segmentation and hybrid multiscale feature extraction for traffic sign detection and recognition. J. Intell. Fuzzy Syst. 1–16 (2018)
Handoko, H., Pratama, J.H., Yohanes, B.W.: Traffic sign detection optimization using color and shape segmentation as pre-processing system. TELKOMNIKA (Telecommunication Computing Electronics and Control) 19(1) (2021)
James, D.: Computer vision based traffic sign sensing for smart transport. J. Innov. Image Process. 1(1), 11–19 (2019)
Liu, W., Lu, X.: Research progress of transformer based on computer vision. Comput. Eng. Sci. 1–17 (2022). http://kns.cnki.net/kcms/detail/11.2127.tp.20211129.1135.004.html
Chen, Q., Hu, Q., Li, J.: Image deblurring method based on dual task convolution neural network. Chin. J. Liquid Cryst. Displays 36(11), 1486–1496 (2021)
Wang, Q.R., Wei, Y.M., Zhu, C.F., et al.: Research on traffic accident risk prediction based on spatio-temporal convolution. Comput. Eng. https://doi.org/10.19678/j.issn.1000-3428.0062961
Yao, R., Xia, S.X., Zhou, Y., et al.: Spatio-temporal graph convolutional networks with attention mechanism for video target segmentation. Chin. J. Graph. 26(10), 2376–2387 (2021)
Jiang, T.T., Liu, Y., Ma, X., et al.: Multi-branch collaborative RGB-T image saliency target detection. J. Image Graph. 26(10), 2388–2399 (2021)
Dai, J., Cao, Y., Shen, Q., et al.: Traffic flow prediction based on multi-spatial-temporal graph convolutional network. Appl. Res. Comput. 1–6 (2022). https://doi.org/10.19734/j.issn.1001-3695.2021.08.0361
Yu, G., Liu, Z., Zhao, P.: A single image collaborative saliency detection method. Appl. Res. Comput. 37(S2), 308–310 (2020)
Lin, H., Li, J., Liang, D., et al.: Complete detection method of salient objects. J. Chin. Comput. Syst. 37(09), 2079–2083 (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Zou, Q., Xiao, L., Xu, G., Wang, X., Mu, N. (2023). Traffic Sign Image Segmentation Algorithm Based on Improved Spatio-Temporal Map Convolution. In: Yu, C., Zhou, J., Song, X., Lu, Z. (eds) Green, Pervasive, and Cloud Computing. GPC 2022. Lecture Notes in Computer Science, vol 13744. Springer, Cham. https://doi.org/10.1007/978-3-031-26118-3_6
Download citation
DOI: https://doi.org/10.1007/978-3-031-26118-3_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-26117-6
Online ISBN: 978-3-031-26118-3
eBook Packages: Computer ScienceComputer Science (R0)