由於出入口及公共場所的人數統計對於安全監控有相當高的參考價值。過去以人工進行人數控管的方式，需要高人力成本。我們利用攝影機取得行人進出的影像，再以電腦視覺處理方式對監視區域內的行人做移動方向之追蹤及行人統計，以取得人流數量資料，未來將可應用於相關之安全監控，而其好處還包括節省人力資源、降低設備成本和未來可以應用於各類型之智慧型安全監控系統。
本系統主要分成三個部分: 1.首先研究三種不同的背景建模演算法實現監控背景之建立及更新，為了使系統能使用於不同監控場景，我們比較了三種背景建模演算法後，我們選擇了處理速度佳，抗雜訊能力好的Codebook方法，作為本論文的背景模型。2.利用Codebook背景模型分離背景與前景目標，將目標物件作連通切割後，取得監視區域範圍內之移動目標區塊。3.為了去除不符合行人之區塊，本論文提出一個行人面積特徵估測法作行人偵測，並將偵測後的各別行人區塊結合運動歷史影像(MHI)之移動追蹤演算法，獲得其行人之運動梯度方向，達成行人移動方向之追蹤及行人計數的目標。
實驗結果顯示，本論文所實現的系統能即時偵測場景中的行人，並追蹤其方向與偵測人數，在執行速度上平均每秒可以處理44 張frame。

The number of pedestrians in public places has very high reference value for security monitoring. In the past, the way to monitor and control the number of people required high labor cost. We use the camera to obtain images of pedestrians, then to track the direction of movement of pedestrians and calculate the number of pedestrians. Finally we get an amount of pedestrians in a particular region. The information can be applied to various types of intelligent safety monitoring system. The benefits include savings personnel costs, reducing equipment costs.
The goal of this thesis is to detect pedestrians. It is divided into three parts. First, the system use image processing technology to establish and update the background images of the surveillance video. Second, the moving object segmentation algorithm is applied to obtain the moving blocks within the scope of surveillance. Then according to the estimation method of area, the non-pedestrian objects are removed. Finally the system use tracking algorithm for pedestrian based on motion history images (MHI) to obtain the gradient direction of its movement, and track the direction of pedestrian and count the numbers of pedestrian.
Experimental results show that the system using codebook can detect pedestrians in the scene and track the pedestrian detection. The average speed in the system is 44 frames per second, so the system can achieve real-time processing.

[1] Yang, D. B. , Gonzalez-Ganos, and L. J. , “Counting people in crowds with a real-time network of simple image,” Proc. Ninth IEEE International Conference on Computer Vision, Vol. 1, pp. 122-129, (2003)
[2] S. Rogic, ”Tracking people using range information,” Available: http://www.cs.ubc.ca/~rogic/projects.html
[3] J. Bescos, J. M. Menendex, and N. Garcia, “DCT based segmentation applied to a scalable zenithal people counter,” IEEE International Conference on Image Processing (ICIP), Vol. 3, pp. 1005-1008, (2003)
[4] I. Haritaoglu, D. Harwood, and L. S. Davis, “Hydra: multiple people detection and tracking using silhouettes,” Proc. International Conference on Image Analysis and Processing, pp. 280-285, (1999)
[5] 羅安鈞，「俯視型行人計數系統」，碩士論文，國立台灣師範大學資訊工程研究所，台北 (2008)。
[6] J. M. Wang, S. W. Chen, S. Cherng, and C. S. Fuh, “People Counting Using Fisheye Camera,” Proc. Of the 20th IPPR Conference on CVGIP, Mauli, Taiwan, (2007)
[7] F. Bartolini, V. Cappellini, and A. Mecocci, “Counting people getting in and out of a bus by real time image-sequence processing,” Image and Vision Computing., Vol. 12 No. 1, pp.36-41, (1994)
[8] A. Albiol, I. Mora, and V. Naranjo, “Real time high density people counter using morphological tools,” IEEE Trans. Intelligent Transportation System, Vol. 2, No. 4, pp. 204-218, (2001)
[9] M. Rossi, and A. Bozzoli, “Tracking and counting moving people,” IEEE International Conference on Image Processing (ICIP), Vol. 3, pp. 212-216, (1994)
[10] G. Sexton, X. Zhang, and G. Redpath, “Advances in automatic counting of pedestrians,” Proc. of the European Convention on Security and Detection, pp. 106-110, (1995)
[11] X. Zhang, and G. Sexton, “Automatic human head location for pedestrian counting,” IEE 6th International Conference on Image Processing and its Applications, pp. 535-540, (1997)
[12] K. Terada, D. Yoshida, S. Oe, and J. Yamaguchi, “A method of counting the passing people by using the stereo images,” IEEE International Conference on Image Processing (ICIP), Vol. 2, pp. 338-342, (1999)
[13] O. Masoud, and N. P. Papanikolopoulos, “A novel method for tracking and counting pedestrians in real-time using a single camera,” IEEE Trans. Vehicular Technology, Vol. 50, No. 5, pp. 1267-1278, (2001)
[14] V. Rabaud, and S. Belongie, “Counting Crowded Moving Objects,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Vol. 1, pp. 705-711 (2006).
[15] Z. Lin, L.S. Davis, D. Doermann, and D. DeMenthon, “Hierarchical Part-Template Matching for Human Detection and Segmentation,” International Conference on Computer Vision, pp. 1-8, (2007).
[16] L. Dong, V. Parameswaran, V. Ramesh, and I. Zoghlami, “Fast Crowd Segmentation Using Shape Indexing,” International Conference on Computer Vision, pp. 1-8, (2007)
[17] A. B. Chan, Z.-S.J. Liang, and N. Vasconcelos, “Privacy preserving crowd monitoring Counting people without people models or tracking,” IEEE Conference on Computer Vision and Pattern Recognition, (2008)
[18] G. K. H. Pang, and C. K. Ng, “Automated people counting using template matching and head search,” Proc. IEEE Conference Mechatronics and Machine Vision in Practice, (2002)
[19] 彭振軒，「使用樣板比對做進出口行人數量統計」，碩士論文，國立中央大學資訊工程系，桃園 (2006)。
[20] 陳志賢，「基於視訊處理的即時人流計數系統之研究」，碩士論文，國立高雄應用科技大學電子工程所，高雄 (2006)。
[21] K. Kim, T. H. Chalidabhongse, D. Harwood, and L. Davis, “Background Modeling and Subtraction by Codebook Construction,” Proc. IEEE International Conference on Image Processing, vol. 5, pp. 3061-3064, (2004)
[22] T. Kohonen, “Improved Versions of Learning Vector Quantization,” International Joint Conference on In Neural Networks, vol. 1, pp. 545-550, (1990)
[23] J. Davis, and A. Bobick, “The Representation And Recognition of Human Movement using Temporal Templates,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 23, pp. 257-267, (2002)
[24] G. R. Bradskil, and J. W. Davis, “Motion Segmentation and Pose Recognition with Motion History Gradients,” Machine Vision and Applications, vol. 13, pp. 174-184, (2001)
[25] 鐘國亮，影像處理與電腦視覺，第三版，東華書局，台北，（2006）。
[26] Rafael C. Gonzalez, and Richard E. Woods, Digital Image Processing, 2nd Edition, Prentice Hall, (2002)
[27] P. KaewTrakulPong, and R. Bowden, “An Improved Adaptive Background Mixture Model for Real-time Tracking with Shadow Detection,” Proc. 2nd European Workshop on Advanced Video Based Surveillance Systems, AVBS01, VIDEO BASED SURVEILLANCE SYSTEMS: Computer Vision and Distributed Processing, (2001)
[28] 鄭士奇，「以高斯模型為基礎並使用陰影消除之動態背景模型建立」，碩士論文，國立交通大學電機與控制工程研究所，新竹 (2005)
[29] The PETS03 Surveillance Ground-Truth Data Sets, International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance (VS-PETS 2003), (2003)