Publication Type : Conference Paper
Publisher : Springer Nature Singapore
Source : Scopus
Url : https://doi.org/10.1007/978-981-19-3035-5_66
Keywords : Accidents; Deep neural networks; Roads and streets; Vehicles; Hybrid approach; LSTM OCR engine (long short-term memory optical character reader); OCR engines; Optical-; Overloaded vehicle detection; Overloaded vehicles; Traffic rules; Traffic violation; Vehicles detection; YOLOv4; Long short-term memory
Campus : Bengaluru
School : School of Computing
Department : Computer Science and Engineering
Year : 2023
Abstract : Nowadays, the number of vehicles on the road is steadily increasing, and this raises the risk of road accidents. One of the reasons of this regrettable situation is violation of traffic rules. Overloaded cars are one example of a traffic rule violation. This research study proposes a hybrid approach for detecting different types of violations and violators. The proposed model will detect the overloaded vehicles and the number plate of the offenders by using the deep neural network model called YOLOv4. Additionally, Tesseract is used to recognize characters in the LSTM number plate. Finally, the proposed model has delivered a satisfactory result. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Cite this Research Publication : Priya Gupta, R. Rajkumar, S. Santhanalakshmi, J. Amudha, Hybrid Approach for Detecting the Traffic Violations Based on Deep Learning Using the Real-Time Data, Scopus, Springer Nature Singapore, 2023, https://doi.org/10.1007/978-981-19-3035-5_66