<div>
	REAL TIME DETECTION AND CLASSIFICATION OF TRAFFIC SIGNS BASED ON YOLO VERSION 3 ALGORITHM  (in English)</div>

Valentin N. Sichkar , Kolyubin Sergey A

2020 , VOLUME 20, NUMBER 3 ( may-june )

ISSN 2226-1494 (print), ISSN 2500-0373 (online)

Publications

Editor-in-Chief

Nikiforov
Vladimir O.
D.Sc., Prof.

Partners

doi: 10.17586/2226-1494-2020-20-3-418-424

REAL TIME DETECTION AND CLASSIFICATION OF TRAFFIC SIGNS BASED ON YOLO VERSION 3 ALGORITHM (in English)

V. N. Sichkar, S. A. Kolyubin

Read the full article

Article in English

For citation:

Sichkar V.N., Kolyubin S.A. Real time detection and classiﬁcation of trafﬁc signs based on YOLO version 3 algorithm. Scientiﬁc and Technical Journal of Information Technologies, Mechanics and Optics, 2020, vol. 20, no. 3, pp. 418–424 (in English). doi: 10.17586/2226-1494-2020-20-3-418-424

Abstract

The issue of effective detection and classiﬁcation of various trafﬁc signs is studied. The two-stage method is proposed for creation of holistic model with end-to-end solution. The ﬁrst stage includes implementation of effective localization of trafﬁc signs by YOLO version 3 algorithm (You Only Look Once). At the ﬁrst stage the trafﬁc signs are grouped into four categories according to their shapes. At the second stage, an accurate classiﬁcation of the located trafﬁc signs is performed into one of the forty-three predeﬁned categories. The second stage is based on another model with one convolutional neural layer. The model for detection of trafﬁc signs was trained on German Trafﬁc Sign Detection Benchmark (GTSDB) with 630 and 111 RGB images for training and validation, respectively. Сlassiﬁcation model was trained on German Trafﬁc Sign Recognition Benchmark (GTSRB) with 66000 RGB images on pure “numpy” library with 19 × 19 dimension of convolutional layer ﬁlters and reached 0.868 accuracy on testing dataset. The experimental results illustrated that the training of the ﬁrst model deep network with only four categories for location of trafﬁc signs produced high mAP (mean Average Precision) accuracy reaching 97.22 %. Additional convolutional layer of the second model applied for ﬁnal classiﬁcation creates efﬁcient entire system. Experiments on processing video ﬁles demonstrated frames per second (FMS) between thirty-six and sixty-one that makes the system feasible for real time applications. The frames per second depended on the number of trafﬁc signs to be detected and classiﬁed in every single frame in the range from six to one.

Keywords: trafﬁc signs detection, deep convolutional neural network, YOLO v3, trafﬁc signs classiﬁcation, detection accuracy

References

Zhu Y., Zhang C., Zhou D., Wang X., Bai X., Liu W. Traffic sign detection and recognition using fully convolutional network guided proposals. Neurocomputing, 2016, vol. 214, pp. 758–766. doi: 10.1016/j.neucom.2016.07.009
Tabernik D., Skocaj D. Deep learning for large-scale traffic-sign detection and recognition. IEEE Transactions on Intelligent Transportation Systems, 2020, vol. 21, no. 4, pp. 1427–1440. doi: 10.1109/TITS.2019.2913588
Chung J.H., Kim D.W., Kang T.K., Lim M.T. Traffic sign recognition in harsh environment using attention based convolutional pooling neural network. Neural Processing Letters, 2020, in press. doi: 10.1007/s11063-020-10211-0
Mehta S., Paunwala C., Vaidya B. CNN based traffic sign classification using adam optimizer. Proc. of the International Conference on Intelligent Computing and Control Systems (ICCS 2019), 2019, pp. 1293–1298. doi: 10.1109/ICCS45141.2019.9065537
Ren S., He K., Girshick R.B., Sun J. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, vol. 39, no. 6, pp. 1137–1149. doi: 10.1109/TPAMI.2016.2577031
Shrivastava A., Gupta A., Girshick R.B. Training region-based object detectors with online hard example mining. Proc. 29^th IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2016), 2016, pp. 761–769. doi: 10.1109/CVPR.2016.89
Zaklouta F., Stanciulescu B. Real-time traffic-sign recognition using tree classifiers. IEEE Transactions on Intelligent Transportation Systems, 2012, vol. 13, no. 4, pp. 1507–1514. doi: 10.1109/TITS.2012.2225618
Ellahyani A., Ansari M.E., Jaafari I.E., Charfi S. Traffic sign detection and recognition using features combination and random forests.International Journal of Advanced Computer Science and Applications, 2016, vol. 7, no. 1, pp. 6861–6931. doi: 10.14569/IJACSA.2016.070193
Zaklouta F., Stanciulescu B. Real-time traffic sign recognition in three stages. Robotics Autonomous Systems, 2014, vol. 62, no. 1, pp. 16–24. doi: 10.1016/j.robot.2012.07.019
Redmon J., Divvala S.K., Girshick R.B., Farhadi A. You only look once: Unified, real-time object detection. Proc. 29^th IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2016), 2016, pp. 779–788. doi: 10.1109/CVPR.2016.91
Redmon J., Farhadi A. YOLO9000: Better, faster, stronger. Proc. 30^th IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2017), 2017, pp. 6517–6525. doi: 10.1109/CVPR.2017.690
Redmon J., Farhadi A. YOLOv3: An incremental improvement. arXiv, 2018, abs/1804.02767.
Houben S., Stallkamp J., Salmen J., Schlipsing M., Igel C. Detection of traffic signs in real-world images: The German traffic sign detection benchmark. Proc. of the International Joint Conference on Neural Networks (IJCNN 2013), 2013, pp. 6706807. doi: 10.1109/IJCNN.2013.6706807
Stallkamp J., Schlipsing M., Salmen J., Igel C. Man vs. computer: Benchmarking machine learning algorithms for traffic sign recognition. Neural Networks, 2012, vol. 32, pp. 323–332. doi: 10.1016/j.neunet.2012.02.016
Sichkar V.N., Kolyubin S.A. Effect of various dimension convolutional layer filters on traffic sign classification accuracy. Scientific and Technical Journal of Information Technologies, Mechanics and Optics, 2019, vol. 19, no. 3, pp. 546–552. doi: 10.17586/2226-1494-2019-19-3-546-552
Davis J., Goadrich M. The relationship between precision-recall and ROC curves. ACM International Conference Proceeding Series, 2006, vol. 148, pp. 233–240. doi: 10.1145/1143844.1143874
Everingham M., Van Gool L., Williams C.K., Winn J.M., Zisserman A. The pascal visual object classes (VOC) challenge. International Journal of Computer Vision, 2010, vol. 88, no. 2, pp. 303–338. doi: 10.1007/s11263-009-0275-4

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License