Small Traffic Sign Detection in Big Images: Searching Needle in a Hay | IEEE Journals & Magazine | IEEE Xplore

Small Traffic Sign Detection in Big Images: Searching Needle in a Hay


Proposed approach for small traffic sign detection. The yellow coloured box indicate proposed anchor box algorithm (sort and scale), whereas blue coloured box represent i...

Abstract:

Traffic sign detection is an essential module of self-driving cars and driver assistance system. The major challenge being, traffic sign appear relatively smaller in road...Show More

Abstract:

Traffic sign detection is an essential module of self-driving cars and driver assistance system. The major challenge being, traffic sign appear relatively smaller in road view images. It covers only 1%-2% of the total image area. Hence, its challenging to detect very small traffic sign in a larger image covering huge background of similar shape objects. Thus, we propose YOLOv3 network layers pruning and patch wise training strategy for small sized traffic sign detection. This aids in improving recall percentage and mean Average Precision. We also propose anchor box selection algorithm that uses bounding box dimension density to obtain optimal anchor set for the dataset. This reduces false positives and log-average miss rate. The proposed approach is evaluated on German traffic sign detection benchmark and Swedish traffic sign dataset and proves that it achieved a good balance between mAP and inference time.
Proposed approach for small traffic sign detection. The yellow coloured box indicate proposed anchor box algorithm (sort and scale), whereas blue coloured box represent i...
Published in: IEEE Access ( Volume: 10)
Page(s): 18667 - 18680
Date of Publication: 10 February 2022
Electronic ISSN: 2169-3536

Funding Agency:

Author image of Yawar Rehman
Department of Electronic Engineering, NED University of Engineering and Technology, Karachi, Sindh, Pakistan
Yawar Rehman received the bachelor’s degree in electronics engineering from the Mehran University of Engineering and Technology, Pakistan, in 2008, and the master’s and Ph.D. degrees in electronics and communication engineering from Hanyang University, South Korea, in 2017. He is currently an Assistant Professor with the NED University of Engineering and Technology, Pakistan. His current research interests include compute...Show More
Yawar Rehman received the bachelor’s degree in electronics engineering from the Mehran University of Engineering and Technology, Pakistan, in 2008, and the master’s and Ph.D. degrees in electronics and communication engineering from Hanyang University, South Korea, in 2017. He is currently an Assistant Professor with the NED University of Engineering and Technology, Pakistan. His current research interests include compute...View more
Author image of Hafsa Amanullah
Dhanani School of Science and Engineering, Habib University, Karachi, Sindh, Pakistan
Hafsa Amanullah received the B.E. degree in electronic engineering and the M.Eng. degree in industrial electronics from the NED University of Engineering and Technology, Pakistan, in 2017 and 2021, respectively.
Currently, she is working as a Senior Research Assistant with Habib University. Her research interests include computer vision, deep learning, and artificial intelligence.
Hafsa Amanullah received the B.E. degree in electronic engineering and the M.Eng. degree in industrial electronics from the NED University of Engineering and Technology, Pakistan, in 2017 and 2021, respectively.
Currently, she is working as a Senior Research Assistant with Habib University. Her research interests include computer vision, deep learning, and artificial intelligence.View more
Author image of Muhammad Ayaz Shirazi
School of Electronics Engineering, IT College, Kyungpook National University, Buk-gu, Daegu, South Korea
Haptics, Human–Robotics and Condition Monitoring Laboratory (National Center of Robotics and Automation), NED University of Engineering and Technology, Karachi, Sindh, Pakistan
Muhammad Ayaz Shirazi received the B.E. degree in electronics engineering from the NED University of Engineering and Technology (NEDUET), Pakistan, in 2012, and the Ph.D. degree from the Optomechatronics and Multi-Scale Robotics Laboratory, Kyungpook National University, South Korea, in 2018. He was the Former Postdoctoral Researcher at KAIST, South Korea. Currently, he is working as a Senior Researcher with NCRA, NEDUET....Show More
Muhammad Ayaz Shirazi received the B.E. degree in electronics engineering from the NED University of Engineering and Technology (NEDUET), Pakistan, in 2012, and the Ph.D. degree from the Optomechatronics and Multi-Scale Robotics Laboratory, Kyungpook National University, South Korea, in 2018. He was the Former Postdoctoral Researcher at KAIST, South Korea. Currently, he is working as a Senior Researcher with NCRA, NEDUET....View more
Author image of Min Young Kim
School of Electronics Engineering, IT College, Kyungpook National University, Buk-gu, Daegu, South Korea
Research Center for Neurosurgical Robotic System, IT College, Kyungpook National University, Buk-gu, Daegu, South Korea
Min Young Kim (Member, IEEE) received the B.S., M.S., and Ph.D. degrees with the Korea Advanced Institute of Science and Technology, South Korea, in 1996, 1998, and 2004, respectively. From 2004 to 2005, he worked as a Senior Researcher at Mirae Corporation. From 2005 to 2009, he worked as a Chief Research Engineer at Kohyoung Corporation, in the research field of artificial vision systems for intelligent machines and rob...Show More
Min Young Kim (Member, IEEE) received the B.S., M.S., and Ph.D. degrees with the Korea Advanced Institute of Science and Technology, South Korea, in 1996, 1998, and 2004, respectively. From 2004 to 2005, he worked as a Senior Researcher at Mirae Corporation. From 2005 to 2009, he worked as a Chief Research Engineer at Kohyoung Corporation, in the research field of artificial vision systems for intelligent machines and rob...View more

Author image of Yawar Rehman
Department of Electronic Engineering, NED University of Engineering and Technology, Karachi, Sindh, Pakistan
Yawar Rehman received the bachelor’s degree in electronics engineering from the Mehran University of Engineering and Technology, Pakistan, in 2008, and the master’s and Ph.D. degrees in electronics and communication engineering from Hanyang University, South Korea, in 2017. He is currently an Assistant Professor with the NED University of Engineering and Technology, Pakistan. His current research interests include computer vision, image processing, feature extraction, and deep learning.
Yawar Rehman received the bachelor’s degree in electronics engineering from the Mehran University of Engineering and Technology, Pakistan, in 2008, and the master’s and Ph.D. degrees in electronics and communication engineering from Hanyang University, South Korea, in 2017. He is currently an Assistant Professor with the NED University of Engineering and Technology, Pakistan. His current research interests include computer vision, image processing, feature extraction, and deep learning.View more
Author image of Hafsa Amanullah
Dhanani School of Science and Engineering, Habib University, Karachi, Sindh, Pakistan
Hafsa Amanullah received the B.E. degree in electronic engineering and the M.Eng. degree in industrial electronics from the NED University of Engineering and Technology, Pakistan, in 2017 and 2021, respectively.
Currently, she is working as a Senior Research Assistant with Habib University. Her research interests include computer vision, deep learning, and artificial intelligence.
Hafsa Amanullah received the B.E. degree in electronic engineering and the M.Eng. degree in industrial electronics from the NED University of Engineering and Technology, Pakistan, in 2017 and 2021, respectively.
Currently, she is working as a Senior Research Assistant with Habib University. Her research interests include computer vision, deep learning, and artificial intelligence.View more
Author image of Muhammad Ayaz Shirazi
School of Electronics Engineering, IT College, Kyungpook National University, Buk-gu, Daegu, South Korea
Haptics, Human–Robotics and Condition Monitoring Laboratory (National Center of Robotics and Automation), NED University of Engineering and Technology, Karachi, Sindh, Pakistan
Muhammad Ayaz Shirazi received the B.E. degree in electronics engineering from the NED University of Engineering and Technology (NEDUET), Pakistan, in 2012, and the Ph.D. degree from the Optomechatronics and Multi-Scale Robotics Laboratory, Kyungpook National University, South Korea, in 2018. He was the Former Postdoctoral Researcher at KAIST, South Korea. Currently, he is working as a Senior Researcher with NCRA, NEDUET. His research interests include camera calibration and 3D reconstruction, digital image processing, augmented reality, deep learning, and structured light imaging.
Muhammad Ayaz Shirazi received the B.E. degree in electronics engineering from the NED University of Engineering and Technology (NEDUET), Pakistan, in 2012, and the Ph.D. degree from the Optomechatronics and Multi-Scale Robotics Laboratory, Kyungpook National University, South Korea, in 2018. He was the Former Postdoctoral Researcher at KAIST, South Korea. Currently, he is working as a Senior Researcher with NCRA, NEDUET. His research interests include camera calibration and 3D reconstruction, digital image processing, augmented reality, deep learning, and structured light imaging.View more
Author image of Min Young Kim
School of Electronics Engineering, IT College, Kyungpook National University, Buk-gu, Daegu, South Korea
Research Center for Neurosurgical Robotic System, IT College, Kyungpook National University, Buk-gu, Daegu, South Korea
Min Young Kim (Member, IEEE) received the B.S., M.S., and Ph.D. degrees with the Korea Advanced Institute of Science and Technology, South Korea, in 1996, 1998, and 2004, respectively. From 2004 to 2005, he worked as a Senior Researcher at Mirae Corporation. From 2005 to 2009, he worked as a Chief Research Engineer at Kohyoung Corporation, in the research field of artificial vision systems for intelligent machines and robots. He was a Visiting Associate Professor with the Department of Electrical and Computer Engineering and the School of Medicine, Johns Hopkins University, from 2014 to 2016. Since 2009, he has been an Assistant Professor with the School of Electrical Engineering and Computer Science, Kyungpook National University. Currently, he is a Full Professor with the School of Electronics Engineering, Kyungpook National University, and the Deputy Director of the Research Center for Neurosurgical Robotic Systems and the KNU-LG Convergence Research Center. His research interests include visual sensor system for robotic perception and recognition, human augmentation devices, control system for microrobotic systems, and surgical robotic systems.
Min Young Kim (Member, IEEE) received the B.S., M.S., and Ph.D. degrees with the Korea Advanced Institute of Science and Technology, South Korea, in 1996, 1998, and 2004, respectively. From 2004 to 2005, he worked as a Senior Researcher at Mirae Corporation. From 2005 to 2009, he worked as a Chief Research Engineer at Kohyoung Corporation, in the research field of artificial vision systems for intelligent machines and robots. He was a Visiting Associate Professor with the Department of Electrical and Computer Engineering and the School of Medicine, Johns Hopkins University, from 2014 to 2016. Since 2009, he has been an Assistant Professor with the School of Electrical Engineering and Computer Science, Kyungpook National University. Currently, he is a Full Professor with the School of Electronics Engineering, Kyungpook National University, and the Deputy Director of the Research Center for Neurosurgical Robotic Systems and the KNU-LG Convergence Research Center. His research interests include visual sensor system for robotic perception and recognition, human augmentation devices, control system for microrobotic systems, and surgical robotic systems.View more

References

References is not available for this document.