這篇研究的目的是想要提供一個高效率多方向輪椅偵測系統。我們使用二種不同特性的特徵來描述輪椅。為了處理輪椅在各種方向視覺上的變化並增進偵測的效能，我們建立了一個基於多方向串接促進器及決策樹的架構的多方向輪以偵測系統。利用這個系統可以很快的去除非輪椅的區域同時偵測輪椅。並能建立一個追蹤系統，去輪椅使用者的照護輪椅使用者的行動。在我們的實驗可以達到百分之九十以上的準確率。

The purpose of this paper is to provide a fast wheelchair detection system. Two features with different-properties are used to represent wheelchairs. To overcome the variable appearances of wheelchairs due to viewing directions and perform a fast detection, a multi-view wheelchair detection system is proposed based on the concepts of the cascade boosting and the decision tree. This system can quickly remove obvious non-wheelchair regions and obtain the moving direction of the detected wheelchair at the same time. When a wheelchair is detected, a tracking procedure is performed to monitor the wheelchair user. Through our experiments, we find that the wheelchair detection rate can achieve over 90 %.

Chapter 1 Introduction ........................................1
Chapter 2 Related Works .......................................4
Chapter 3 Multi-Directional Wheelchair Detection Framework ....7
3.1 Decision tree structure ...................................7
3.2 Boosted classifier as Tree node...........................10
3.2.1 Boosted cascade classifier..............................11
3.2.2 Feature pool creation...................................15
3.3 Training implementation...................................19
3.3.1 Feature creation........................................20
3.3.2 Training data of cascade................................21
3.3.3 Train boosted cascade classifiers ......................21
3.4 Detection process ........................................23
3.5 Multi-directional tracking ...............................24
Chapter 4 Experimental Results................................29
4.1 The results of different features with each direction.....29
4.2 The result of multi-directional detection.................35
4.3 The result of multi-directional tracking .................37
Chapter 5 Conclusion..........................................38
References....................................................39

[1] Paul Viola, Michael Jones, "Rapid object detection using a boosted cascade of simple features", Proc. IEEE Conf. Computer Vision and Pattern Recognition, vol.1, pp. 511-518, 2001
[2] Paul Viola, Michael Jones, "Robust Real Time Object Detection",Second International Workshop on Statistical and Computational Theories of Vision, 2002
[3] Bo WU, Haizhou AI, Chang HUANG, Shihong LAO, "Fast Rotation Invariant Multi-View Face Detection Based on Real Adaboost", Pro. Sixth IEEE Int'l Conf. Automatic Face and Gesture Recognition, pp. 79- 84, 2004
[4] Ashish Myles, Dr. Niels Da Vitoria Lobo, Dr. Mubarak Shah, "Wheelchair Detection in a Calibrated Environment", The 5th Asian Conf. Computer Vision, 2002
[5] Paul Beardsley, "Wheelchair detection using stereo vision", MERL Technical report, 2002
[6] Bergasa, L.M. Mazo, M. Gardel, A. Garcia, J.C. Ortuno, A. Mendez, A.E., "Guidance of a wheelchair for handicapped people by face tracking", Proc. 7th IEEE Int'l Conf. Emerging Technologies and Factory Automation, vol. 1, pp. 105-111, 1999
[7] Cyril Cauchois, Fabrice de Chaumont, Bruno Marhic , Laurent Delahoche, Mélanie Delafosse, "Robotic assistance: an automatic wheelchair tracking and following functionality by omnidirectional vision", IEEE/RSJ Int'l Conf. Intelligent Robots and Systems, pp. 2560-2565, 2005
[8] Datong Chen, Jie Yang, Howard D. Wactlar, "Towards automatic analysis of social interaction patterns in a nursing home environment from video", International Multimedia Conference Proceedings of the 6th ACM SIGMM international workshop on Multimedia information retrieval, pp. 283-290, 2004
[9] Chin-Ann Yang, Pau-Choo Chung, "Recovery of 3-D location and orientation of a wheelchair in a calibrated environment by using single perspective geometry", IEEE TENCON, pp. 1-4, 2007
[10] Rainer Lienhart, Jochen Maydt, "An Extended Set of Haar-like Features for Rapid Object Detection", Proc. IEEE Int'l Conf. Image Processing, vol.1, pp. 900-903, 2002
[11] Rainer Lienhart, Alexander Kuranov, Vadim Pisarevsky, "Empirical Analysis of Detection Cascades of Boosted Classifiers for Rapid Object Detection", MRL Technical Report, 2002
[12] Navneet Dalal, Bill Triggs, "Histograms of Oriented Gradients for Human Detection," Proc. IEEE Conf. Computer Vision and Pattern Recognition, vol. 1, pp. 886-893, 2005
[13] Qiang Zhu, Shai Avidan, Mei-Chen Yeh, Kwang-Ting Cheng, "Fast Human Detection Using a Cascade of Histograms of Oriented Gradients," Proc. IEEE Conf. Computer Vision and Pattern Recognition, vol. 2, pp. 1491 - 1498, 2006
[14] Chi-Chen Raxle Wang, Jenn-Jier James Lien, "AdaBoost Learning for Human Detection Based on Histograms of Oriented Gradients," The 8th Asian Conf. Computer Vision, vol. 4843, pp. 885-895, 2007
[15] Yu-Ting Chen, Chu-Song Chen, "A Cascade of Feed-Forward Classifiers for Fast Pedestrian Detection," The 8th Asian Conf. Computer Vision, vol. 4843, pp. 905-914, 2007
[16] Fatih Porikli, "Integral histogram: a fast way to extract histograms in Cartesian spaces," Proc. IEEE Conf. Computer Vision and Pattern Recognition, vol. 1, pp. 829- 836, 2005
[17] Yu-Ting Chen, Chu-Song Chen, "Cascading Rectangle and Edge Orientation Features for Fast Pedestrian Detection," Third Int’l Conf. on Intelligent Information Hiding and Multimedia Signal Processing, vol. 2, pp. 407-410, 2007
[18] Chun-Rong Huang, Chu-Song Chen, Pau-Choo Chung, "Contrast Context Histogram - An Efficient Discriminating Local Descriptor for Object Recognition and Image Matching," Pattern Recognition, vol. 41, no. 10, pp. 3071-3077, 2008.
[19] Chun-Rong Huang, Chu-Song Chen, Pau-Choo Chung, "Contrast Context Histogram - A Discriminating Local Descriptor for Image Matching," Proc. IEEE Int'l Conf. Pattern Recognition, vol. 4, pp. 53-56, 2006
[20] Robert E. Schapire, Yoram Singer, "Improved Boosting Algorithms Using Confidence-rated Predictions", Machine Learning, vol. 37, no. 3, pp. 297-336, 1999
[21] Yoav Freund, Robert E. Schapire, "Experiments with a New Boosting Algorithm", Proc. Thirteenth Int'l Conf. Machine Learning, 1996
[22] David G. Lowe, "Distinctive Image Features from Scale-Invariant Keypoints," Int'l J. of Computer Vision, vol. 60, no. 2, pp. 91-110, 2004
[23] David G. Lowe, "Object Recognition from Local Scale-Invariant Features," Seven IEEE Int'l Conf. Computer Vision, vol. 2, pp. 1150-1157, 1999
[24] Kyungnam Kim, Thanarat H. Chalidabhongse, David Harwood, Larry Davis, "Background modeling and subtraction by codebook construction", Proc. IEEE Int'l Conf. Image Processing, vol. 5, pp. 3061-3064, 2004