Abstract
Automatic detection of road safety attributes is an important step in designing a reliable road safety system. Due to the outstanding performance over the handcraft feature extraction-based methods for detecting objects, deep learning can be used to develop a robust road safety system. However, there are many challenges in using deep learning models. Firstly, they require a large dataset for training. Secondly, a class imbalance is a common problem in deep learning models. Finally, when a new attribute is introduced to a deep learning model, the whole model must be re-trained using all training samples which requires a lot of time. In order to solve some of these problems, we propose a novel single class detection-based deep learning approach for the identification of safety attributes in roadside video data. The approach is based on fusion of multiple fully convolutional network (FCN) models. Each model is trained to detect a single attribute/class using two classes (single attribute vs all other attributes) datasets. The proposed approach was evaluated on data provided by the Department of Transport and Main Roads (DTMR). The proposed approach achieved high accuracy and a new attribute can be added to the system without retraining the whole system.
Similar content being viewed by others
References
Road Safety Statistics, Queensland Government, Department of Transport and Main Roads. [Online]. http://www.tmr.qld.gov.au/Safety/Transport-and-road-statistics/Road-safety-statistics.aspx. Accessed 9 Dec 2019
Luo X, Zhu J, Yu Q (2019) Efficient convNets for fast traffic sign recognition. IET Intell Transp Syst 13(6):1011–1015
Wang C (2018) Research and application of traffic sign detection and recognition based on deep learning. In: International conference robots and intelligent systems (ICRIS), Changsha, China, pp 150–152
Hur J, Kang S-N, Seo S-W (2013) Multi-lane detection in urban driving environments using conditional random fields. In: Proceedings on IEEE intelligent vehicle symposium (IV), Gold Coast, Queensland, Australia, pp 1297–1302
Wu P-C, Chang C-Y, Lin CH (2014) Lane-mark extraction for automobiles under complex conditions. Pattern Recognit 47(8):2756–2767
Yang Y, Luo H, Xu H, Wu F (2016) Towards real-time traffic sign detection and classification. IEEE Trans Intell Transp Syst (ITS) 17(7):2022–2031
Gonzalez Á et al (2011) Automatic traffic signs and panels inspection system using computer vision. IEEE Trans Intell Transp Syst (ITS) 12(2):485–499
Barnes N, Zelinsky A, Fletcher LS (2008) Real-time speed sign detection using the radial symmetry detector. IEEE Trans Intell Transp Syst (ITS) 9(2):322–332
Hoang TM, Nam SH, Park KR (2019) Enhanced detection and recognition of road markings based on adaptive region of interest and deep learning. IEEE Access 7:109817–109832
Lee S et al (2017) VPGNet: vanishing point guided network for lane and road marking detection and recognition. In: Proceedings on IEEE international conference on computer vision (ICCV), Venice, Italy, pp 1965–1973
Acilo JPN, Cruz AGSD, Kaw MKL, Mabanta MD, Pineda VGG, Roxas EA (2018) Traffic sign integrity analysis using deep learning. In: IEEE 14th international colloquium on signal processing and its applications (CSPA), Batu Feringghi, Malaysia, pp 107–112
Li J, Mei X, Prokhorov D, Tao D (2017) Deep neural network for structural prediction and lane detection in traffic scene. IEEE Trans Neural Netw Learn Syst 28(3):690–703
Cho SJ, Seong Kim B, Kim TS, Kong S (2019) Enhancing GNSS performance and detection of road crossing in urban area using deep learning. In: IEEE intelligent transportation systems conference (ITSC), Auckland, New Zealand, pp 2115–2120
Aly M (2008) Real time detection of lane markers in urban streets. In: IEEE intelligent vehicles symposium (IV), Eindhoven, Netherlands, pp 7–12
Bangquan X, Xiong WX (2019) Real-time embedded traffic sign recognition using efficient convolutional neural network. IEEE Access 7:53330–53346
Zhang W, Mi Z, Zheng Y, Gao Q, Li W (2019) Road marking segmentation based on siamese attention module and maximum stable external region. IEEE Access 7:143710–143720
Yuan Y, Xiong Z, Wang Q (2017) An incremental framework for video-based traffic sign detection, tracking, and recognition. IEEE Trans Intell Transp Syst (ITS) 18(7):1918–1929
He B, Ai R, Yan Y, Lang X (2016) Accurate and robust lane detection based on dual-view convolutional neutral network. In: Proceedings on IEEE intelligent vehicles symposium (IV), Gothenburg, Sweden, pp 1041–1046
Chen T, Chen Z, Shi Q, Huang X (2015) Road marking detection and classification using machine learning algorithms. In: Proceedings on IEEE intelligent vehicles symposium (IV), Seoul, South Korea, pp 617–621
Lee HS, Kim K (2018) Simultaneous traffic sign detection and boundary estimation using convolutional neural network. IEEE Trans Intell Transp Syst (ITS) 19(5):1652–1663
Zou Q, Zhang Z, Li Q, Qi X, Wang Q, Wang S (2019) DeepCrack: learning hierarchical convolutional features for crack detection. IEEE Trans Image Process 28(3):1498–1512
Mandal V, Uong L, Adu-Gyamfi Y (2018) Automated road crack detection using deep convolutional neural networks. In: IEEE international conference on big data (Big Data), Seattle, WA, USA, pp 5212–5215
Jenkins MD, Carr TA, Iglesias MI, Buggy T, Morison G (2018) A deep convolutional neural network for semantic pixel-wise segmentation of road and pavement surface cracks. In: 26th European signal processing conference (EUSIPCO), Rome, pp 2120–2124
Carr TA, Jenkins MD, Iglesias MI, Buggy T. Morison G (2018) Road crack detection using a single stage detector based deep neural network. In: IEEE workshop environmental, energy, and structural monitoring systems (EESMS), Salerno, Italy, pp 1–5
Hoang TM, Nguyen PH, Truong NQ, Lee YW, Park KR (2019) Deep retinaNet-based detection and classification of road markings by visible light camera sensors. Sensors 19:281
Zhu Y, Liao M, Yang M, Liu W (2018) Cascaded segmentation-detection networks for text-based traffic sign detection. IEEE Trans Intell Transp Syst (ITS) 19(1):209–219
Wu T, Ranganathan A (2012) A practical system for road marking detection and recognition. In: IEEE intelligent vehicles symposium (IV), Alcala de Henares, Spain, pp 25–30
Jan Z, Verma B, Affum J, Atabak S, Moir L (2018) A convolutional neural network based deep learning technique for identifying road attributes. In: International conference on image and vision computing New Zealand (IVCNZ), Auckland, New Zealand, pp 1–6
Sanjeewani TGP, Verma B (2019) Learning and analysis of AusRAP attributes from digital video recording for road safety. In: International conference on image and vision computing New Zealand (IVCNZ), Dunedin, New Zealand, pp 1–6
Cambridge-Driving Labeled Video Database (CamVid). http://mi.eng.cam.ac.uk/research/projects/VideoRec/CamVid/. Accessed 10 Nov 2020
Daimler Urban Segmentation Dataset. http://www.6d-vision.com/scene-labeling. Accessed 10 Nov 2020
The Málaga Stereo and Laser Urban Data Set—MRPT. https://www.mrpt.org/MalagaUrbanDataset. Accessed 10 Nov 2020
Funding
This research was funded by Australian Research Council (ARC)’s Linkage Projects funding scheme (Grant Number LP170101255).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
Pubudu Sanjeewani is receiving PhD stipend from ARC linkage project (grant number LP170101255) and Professor Verma is a chief investigator for this ARC linkage project.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Sanjeewani, P., Verma, B. Single class detection-based deep learning approach for identification of road safety attributes. Neural Comput & Applic 33, 9691–9702 (2021). https://doi.org/10.1007/s00521-021-05734-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-021-05734-z