An Intelligent Automatic Number Plate Recognition System Based on Computer Vision and Edge Computing
2022 International Conference on Smart Information Systems and Technologies, SIST 2022, Nur-Sultan, Kazakistan, 28 - 30 Nisan 2022, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/sist54437.2022.9945787
- Basıldığı Şehir: Nur-Sultan
- Basıldığı Ülke: Kazakistan
- Anahtar Kelimeler: edge computing, image processing, internet of things, network technology, raspberry pi, video analytic
- Ankara Yıldırım Beyazıt Üniversitesi Adresli: Evet
Özet
© 2022 IEEE.The current traffic monitoring system accumulates a large amount of data and apply algorithms to process the received photo and video images through cameras. These data are used to regulate traffic flows, recognize number plate and record traffic collisions. The collected data are traditionally sent to data centers for processing, analyzing, and then sent to the operator. Thus, a large amount of incoming data to the data center and to the traffic monitoring center reduces the network bandwidth and has a negative impact on the speed of decision-making by the operator. To solve this problem, we proposed to use the edge computing paradigm by placing edge nodes in the near vicinity of the cameras and conducting data preprocessing on edge devices. This paper considers the Raspberry Pi as an edge device, as well as computer vision algorithms for pre-processing data at the edge device and the process of transferring data to the data center with the following prompt decision-making by the operator. As a result of usage the edge computing paradigm helps to optimize network traffic and offload the computing power of the data center.