Computer vision based road traffic accident and anomaly detection in the context of Bangladesh

Author

Lutfe Elahi, Mirza M. ; Yasir, Rahat ; Syrus, Minhaz Ahmed ; Nine, Md S. Q. Zulkar ; Hossain, Ismail ; Ahmed, Nova

Author_Institution

Dept. of Electr. & Comput. Eng., North South Univ., Dhaka, Bangladesh

fYear

2014

fDate

23-24 May 2014

Firstpage

1

Lastpage

6

Abstract

Road traffic accident is one of the major causes of death in Bangladesh. In this article, we propose a method that uses road side video data to learn the traffic pattern and uses vision based techniques to track and determine various kinetic features of vehicles. Finally, the proposed method detects anomaly (possibility of accident) and accidents on the road. The proposed method shows approximately 85% accuracy level in detecting special situations.

Keywords

computer vision; road accidents; road traffic; video signal processing; Bangladesh; computer vision based anomaly detection; computer vision based road traffic accident; kinetic features; road side video data; traffic pattern; Accidents; Feature extraction; Neurons; Probabilistic logic; Roads; Vectors; Vehicles; Bangladesh; computer vision; feature extraction; machine learning; neural network; road traffic accident detection; vehicle detection;

fLanguage

English

Publisher

ieee

Conference_Titel

Informatics, Electronics & Vision (ICIEV), 2014 International Conference on

Conference_Location

Dhaka

Print_ISBN

978-1-4799-5179-6

Type

conf

DOI

10.1109/ICIEV.2014.6850780

Filename

6850780