Automatic traffic surveillance system for vehicle tracking and classification

DMCA

Automatic traffic surveillance system for vehicle tracking and classification (2006)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Junwei Hsieh , Shih-hao Yu , Yung-sheng Chen , Wen-fong Hu , An Automatic , Jun-wei Hsieh , Shih-hao Yu , Yung-sheng Chen , Wen-fong Hu

Venue:	Proc. IEEE Intelligent Transportation Systems Conference
Citations:	31 - 0 self

BibTeX

@INPROCEEDINGS{Hsieh06automatictraffic,
    author = {Junwei Hsieh and Shih-hao Yu and Yung-sheng Chen and Wen-fong Hu and An Automatic and Jun-wei Hsieh and Shih-hao Yu and Yung-sheng Chen and Wen-fong Hu},
    title = {Automatic traffic surveillance system for vehicle tracking and classification},
    booktitle = {Proc. IEEE Intelligent Transportation Systems Conference},
    year = {2006},
    pages = {175--187}
}

OpenURL

Abstract

This paper presents an automatic traffic surveillance system to estimate important traffic parameters from video sequences using only one camera. Different from traditional methods which can classify vehicles to only cars and non-cars, the proposed method has a good capability to categorize vehicles into more specific classes by introducing a new “linearity” feature in vehicle representation. In addition, the proposed system can well tackle the problem of vehicle occlusions caused by shadows, which often lead to the failure of further vehicle counting and classification. This problem is solved by taking advantages of a line-based shadow algorithm which uses a set of lines to eliminate all unwanted shadows. The used lines are devised from the information of different lane dividing lines. Therefore, an automatic scheme to detect lane dividing lines is also proposed. The found lane dividing lines also can provide important information for feature normalization, which can make the vehicle size more invariant and thus much enhance the accuracy of vehicle classification. Once all features are extracted, an optimal classifier is then designed to robustly categorize vehicles into different classes. When recognizing a vehicle, the designed classifier can collect different evidences from its trajectories and the database to make the best decision for vehicle classification. Since more evidences are used, more robustness of classification can be achieved. Experimental results show that the proposed method is more robust, accurate, and powerful than other traditional methods, which utilize only vehicle size and single frame for classification.

Keyphrases

automatic traffic surveillance system vehicle tracking vehicle size traditional method vehicle classification feature normalization video sequence important traffic parameter specific class used line new linearity feature good capability line-based shadow algorithm vehicle representation different lane unwanted shadow found lane vehicle counting different evidence vehicle occlusion different class optimal classifier important information single frame experimental result automatic scheme

DMCA