Qing LIN, Young-joon HAN, Hern-soo HAHN
Department of Electronic Engineering, Soongsil University, Seoul 156-743, South Korea
Abstract-Vehicle detection in still images is a comparatively difficult task. This paper presents a method for this task using boosted local p attern detector constructed from two local features including Haar-like and ori ented gradient features. The whole process is composed of three stages. In the f irst stage, local appearance features of vehicle and non-vehicle objects are ex tracted. Haar-like and oriented gradient features are extracted separately in t his stage as local features. For the second stage, Adaboost algorithm is used to select the most discriminative features as weak detectors from the two local fe ature sets, and a strong local pattern detector is built by the weighted combina tion of these selected weak detectors. Finally, vehicle detection can be perfor med on still images by using boosted strong local feature detector. Experiment r esults show that the local pattern detector constructed in this way combines the advantages of Haar-like and oriented gradient features, and can achieve better detection results than the detector using single Haar-like features.
Key words-vehicle detection; still image; Adaboost; loc al features
Manuscript Number: 1674-8042(2010)01-0041-05
dio: 10.3969/j.issn.1674-8042.2010.01.08
References
[1]C. Pang, W. Lam, N. Yung, 2004. A novel method for resoving vehilce occlusion in a monocular traffic-image sequence. IEEE Transactions on Intelligent Transprotation Systems, 5(3): 129-141.
[2]X. Song, R. Nevatia, 2005. A Model-Based Vehicle Segmentation Metho d for Tracking. Proc. IEEE International Conference on Computer Vision (ICCV05), p. 1124-1131.
[3]Y. Jia, C. Zhang, 2009. Front-view vehicle detection by markov chai n monte carlo method. Pattern Recognition, 42(3): 313-321.
[4]J. Wu, X. Zhang, J. Zhou, 2001. Vehicle Detection in Static Road Ima ges with PCA-and-Wavelet-Based Classifier. Proceedings of the Intelligent Tra nsportation Systems, p.740-744.
[5]L. Tsai, J. Hsieh, K. Fan, 2005. Vehicle Detection using Normalized Color and Edge Map. IEEE Trans. on Image Porcessing, (ICIP 05),, 16(3): 850-864.
[6]Z. Sun, G. Bebis, R. Miller, 2005. On-road Vehicle Detection using Gabor Flters and Support Vector Machines. IEEE Internanational Conference on Dig ital Signal Processing, (ICDSP02), p.125-137.
[7]P. Viola, M. Jones, 2001. Rapid Object Detection Using a Boosted Cas cade of Simple Features. Proc. IEEE Computer Society Conference on Computer Visi on and Pattern Recognition (ICCVPR 01), p.511-518.
[8]C. Papageorgiou, M. Oren, T. Poggio, 1998. A General Framework for O bject Detection. International Conference on Computer Vision, p.555-562.
[9]N. Dalal, B. Triggs, 2005. Histograms of Oriented Gradients for Huma n Detection. Proc. CVPR(05), p. 886-893.
[10]X. Wen, H. Yuan, C. Yang, et al, 2007. Improved Haar Wavelet Featur e Extraction Approaches for Vehicle Detection. IEEE Intelligent Transportation S ystems Conference, IEEE Press, p.1050-1053.
[11]R. Lienhart, J. Maydt, 2002. An extended set of haar-like features for rapid object detection. Proceedings of International Conference on Image Pr ocessing(ICIP02), p. 900-903.
[12]Yotam Abramson, Bruno Steux, Hicham Ghorayeb, 2005. YEF (Yet Even F aster) Real-time object detection. Proceedings of International Workshop on Au tomatic Learning and Real-Time, p.102-112.
[13]P. Sabzmeydani, G. Mori, 2007. Detecting Pedestrians by Learning Sh apelet Features. IEEE Conference on Computer Vision and Pattern Recognition (ICV PR07), p. 1-8.
[14]K. Levi, Y. Weiss, 2004. Learning Object Detection from a Small Num ber of Examples: the Importance of Good Features. Proc. of the IEEE Conference o n CVPR2004, p. 53-60.
[15]P. A. Viola, M. J. Jones, 2004. Robust real-time face detection. Internation Journal of Computer Vision, 57(2): 137-154.
[16]R. E. Schapire, Y. Singer, 1999. Improved boosting algorithms using confidence-Rated Predictions. Machine Learning, 37(3): 297 -336.
[17]D. Withopt, B. Jahne, 2007. Improved Training Algorithm for Tree-l ike Classifiers and its Application to Vehicle Detection. Proc. IEEE Internation al Conference on Intelligent Transportation Systems (ITSC07), p. 642-647.
[full text view]
