Development of an Electric Scooter Photo Recognition System Using YOLO
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The use of electric scooter (e-scooter) sharing services has increased significantly in recent years due to theirconvenience and economy. In order to rent an e-scooter, a user first finds nearby e-scooters using a smartphoneapplication, which shows the global positioning system (GPS) locations of e-scooters around the user. However,since the error of GPS can be more than 10 m, the user may have difficulty finding the exact location of the escooterthe user wants to use. To alleviate this problem, an e-scooter sharing service “Kickgoing,” operated byOlulo in South Korea, provides users with e-scooter photos taken by users upon return, along with their GPSlocations, on its smartphone application. Those photos help subsequent users to find e-scooters more accurately. However, since some users upload photos that do not include e-scooters or are unrecognizable, it is essential toprovide users with only those photos that clearly include an e-scooter. Therefore, in this paper, we develop ane-scooter photo recognition system that can accurately recognize only those photos that include e-scooters. Thedeveloped system, which is based on YOLO, uses three techniques: if a whole e-scooter is not recognized, itrecognizes an e-scooter by recognizing its parts individually; it recognizes e-scooters with significantlydifferent photography angles as different classes; and it provides users with only those photos in which theproportion of the e-scooter is within a certain range. Experimental results on a real dataset show that thedeveloped system recognizes e-scooter photos more accurately compared to a system that uses the YOLOmodel as is.
Abstract
The use of electric scooter (e-scooter) sharing services has increased significantly in recent years due to theirconvenience and economy. In order to rent an e-scooter, a user first finds nearby e-scooters using a smartphoneapplication, which shows the global positioning system (GPS) locations of e-scooters around the user. However,since the error of GPS can be more than 10 m, the user may have difficulty finding the exact location of the escooterthe user wants to use. To alleviate this problem, an e-scooter sharing service “Kickgoing,” operated byOlulo in South Korea, provides users with e-scooter photos taken by users upon return, along with their GPSlocations, on its smartphone application. Those photos help subsequent users to find e-scooters more accurately. However, since some users upload photos that do not include e-scooters or are unrecognizable, it is essential toprovide users with only those photos that clearly include an e-scooter. Therefore, in this paper, we develop ane-scooter photo recognition system that can accurately recognize only those photos that include e-scooters. Thedeveloped system, which is based on YOLO, uses three techniques: if a whole e-scooter is not recognized, itrecognizes an e-scooter by recognizing its parts individually; it recognizes e-scooters with significantlydifferent photography angles as different classes; and it provides users with only those photos in which theproportion of the e-scooter is within a certain range. Experimental results on a real dataset show that thedeveloped system recognizes e-scooter photos more accurately compared to a system that uses the YOLOmodel as is.
참고문헌 (13)
- S. J. Kim, "Study on shared e-scooter usage characteristics and influencing factors", The Journal of The Korea Institute of Intelligent Transport Systems, 20(1), pp.40~53 (2021)[DOI]
- "Kickgoing"[URL]
- "E-scooter rider detection and classification in dense urban environments" (2022)[URL]
- P. Jiang, "A review of Yolo algorithm developments", Procedia Computer Science, 199, pp.1066~1073 (2022)[DOI]
- Z. Q. Zhao, "Object detection with deep learning : a review", IEEE Transactions on Neural Networks and Learning Systems, 30(11), pp.3212~3232 (2019)[DOI]
- "Faster R-CNN: towards real-time object detection with region proposal networks" (2015)[URL]
- R. Girshick, "Rich feature hierarchies for accurate object detection and semantic segmentation", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.580~587 (2014)
- R. Girshick, "Fast R-CNN", Proceedings of 2015 IEEE International Conference on Computer Vision (ICCV), pp.1440~1448 (2015)
- A. Aralikatti, "Real-time object detection and face recognition system to assist the visually impaired", Journal of Physics: Conference Series, 1706(1) (2020)[DOI]
- T. Diwan, "Object detection using YOLO : challenges, architectural successors, datasets and applications", Multimedia Tools and Applications, 82(6), pp.9243~9275 (2023)[DOI]
- C. Kim, "Development of a YOLO-based electric kick scooter photo recognition system", Proceedings of Annual Conference of the Korea Information Processing Society (ACK), pp.622~624 (2022)
- H. Kim, "Development of a real-time automatic passenger counting system using head detection based on deep learning", Journal of Information Processing Systems, 18(3), pp.428~442 (2022)[DOI]
- "YOLO5" (2020)[URL]