High Occupancy Vehicle Detection
High occupancy vehicles lanes (HOV) are highway lanes usually reserved to vehicles carrying at least two persons. They are designed to help move more people though congested areas. In this context, automatic passenger counting systems could be useful to grant access to or to control vehicles in those lanes.
In this work, we propose a real-time passenger detection system based on the analysis of visual images. Each person is detected by mixing the information from different types of classifiers in order to make the detection process faster and more robust.
KeywordsFalse Detection Safety Belt Congested Area Highway Lane Rapid Object Detection
- 1.Billheimer, J.W., Kaylor, K., Shade, C.: Use of videotape in hov lane surveillance and enforcemente. Technical Report 6429, California Department of Transportation (1990)Google Scholar
- 3.Video-Occupancy, C.: Cyclops, occupancy counter system (2005), http://www.vehicleoccupancy.com/products.htm
- 4.Schijns, S., Matthews, P.: A breakthrough in automated vehicle occupancy monitoring sustems for hov/hot facilities. In: 12th HOV Systems Conference (2005)Google Scholar
- 5.Viola, P., Jones, M.: An extended set of haar-like features for rapid object detection. In: IEEE ICIP, vol. 1, pp. 900–903 (2002)Google Scholar
- 6.Lienhart, R., Maydt, J.: Rapid object detection using a boosted cascade of simple features. In: IEEE Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 511–518 (2001)Google Scholar
- 8.Friedman, J., Bentley, J., Finkel, R.: An algorithm for finding best matches in logarithmic expected time. In: ACM Transactions on Mathemathical Software, vol. 3, pp. 209–226 (1977)Google Scholar
- 9.Davies, B., Bornet, O., Bradki, G.: altera: Open Computer Vision Library. Version 1.0 (2006), http://sourceforge.net/projects/opencvlibrary/