Analysis of Object Detection Methods to Detect Traffic Flow

Traffic analysis has received more interest as smart cities become a reality. A key component of traffic analysis is detecting the amount of cars that pass certain points. However, limited research exists that explores methods for the car detection component in these systems. This paper will discuss different computer vision methods that can be used for the detection and analysis of vehicles on the road for active traffic flow analysis and implements them in an experiment to find the best method for the task of car detection. MobileNet and Haar-Cascade based methods are implemented and a compared according to performance and accuracy levels in real-world scenarios. Lastly, the results achieved from the experimental model will be discussed giving detail to why Haar cascade gives better performance and accuracy in most scenarios with an average frame rate of over 40 fps on HD video.

[1]  Mubarak Shah,et al.  Learning object motion patterns for anomaly detection and improved object detection , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[2]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[3]  Katsushi Ikeuchi,et al.  Traffic monitoring and accident detection at intersections , 2000, IEEE Trans. Intell. Transp. Syst..

[4]  Lawrence A. Klein.  ITS Sensors and Architectures for Traffic Management and Connected Vehicles , 2017 .

[5]  Franz J. Meyer,et al.  Traffic monitoring with spaceborne SAR - Theory, simulations, and experiments , 2007, Comput. Vis. Image Underst..

[6]  Heinrich H Bülthoff,et al.  Image-based object recognition in man, monkey and machine , 1998, Cognition.

[7]  Fei-Yue Wang,et al.  Traffic Flow Prediction With Big Data: A Deep Learning Approach , 2015, IEEE Transactions on Intelligent Transportation Systems.

[8]  Josep Perarnau,et al.  Traffic Simulation with Aimsun , 2010 .

[9]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[10]  Bo Li,et al.  A review on vision-based pedestrian detection in intelligent transportation systems , 2012, Proceedings of 2012 9th IEEE International Conference on Networking, Sensing and Control.

[11]  Eugenio Culurciello,et al.  An Analysis of Deep Neural Network Models for Practical Applications , 2016, ArXiv.

[12]  Bo Chen,et al.  MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications , 2017, ArXiv.