利用多相機之動態物件追蹤｜國立清華大學博碩士論文庫

簡易檢索 / 詳目顯示

回結果列表

研究生：	孫翊庭 Yi-Ting Sun
論文名稱：	利用多相機之動態物件追蹤 Dynamic Object Tracking Using Multiple Cameras
指導教授：	黃仲陵 Chung-Lin Huang
口試委員:
學位類別：	碩士 Master
系所名稱：	電機資訊學院 - 電機工程學系 Department of Electrical Engineering
論文出版年：	2007
畢業學年度：	96
語文別：	中文
論文頁數：	41
中文關鍵詞：	物件追蹤、樣板比對
外文關鍵詞：	Object tracking, template matching
相關次數：	點閱：3 下載：0
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

本論文是針對戶外物體的動態追蹤進行研究，我們提出了一套使用兩台PTZ相機的追蹤系統，一台PTZ相機負責全景監控，另一台PTZ相機則負責物件之細部追蹤。本系統會依據所偵測到目標物的大小與速度，分別使用不同的追蹤演算法，使得我們可以追蹤室外廣大區域的人物或車輛目標。我們利用:(1)樣板比對與預測物體行動方向及速度的方式，於畫面中快速掃瞄目標所在位置；可在廣大區域中，追蹤極小之人物目標；(2)利用數學推導，計算出相機經過pan/tilt後，物體座標的移動軌跡，利用該軌跡數學式，我們可即時追蹤快速移動之車體目標。在追蹤人物與車輛的過程中，利用目標物移動所造成的前景差異，可得到目標物此時在畫面中所顯示的大小，藉以調整適當之放大/縮小倍率，以維持追蹤物件的顯示大小。在樣板比對過程，人物的移動速度、上一時刻位置與人物的移動方向…等資訊都被紀錄，作為下一次樣板比對順序的參考；在每一次的樣板比對過程中，若單一色彩錯誤率或是樣板面積比對錯誤率超過臨界值，該畫面區域即被跳過；藉由前述步驟，我們可以大幅縮短樣板比對時間，達到即時追蹤的目的。另一方面，利用樣板比對，我們亦可追蹤被長時間遮蔽之目標物件。論文實驗中，我們針對不同的室外場景、目標長時間遮蔽、車輛與人物快速移動….等不同的狀況進行目標追蹤，驗證系統效能。

The detailed and accurate outdoor object tracking is an important task for surveillance system. In this thesis, we proposed a tracking system to track human or vehicle by two PTZ cameras. For the vehicle tracking purpose, we find the relationship between the cameras and studied the trajectory equation of certain point in the image plane during the panning or tilting procedure of the PTZ camera. We can use the moving trajectory to get the required pan/tilt value for displaying the target in the center of the image plane. In the other hand, an effective template matching method is proposed for human tracking. By considering the information of the moving speed and moving direction of the target, we can track the target instantly even long occlusion situation occurred.

CONTENTS
CHAPTER 1  INTRODUCTION……………………………..………1
1.1    Motivation  1
1.2    Related Works  2
1.3    System Overview  3
1.4    Organization of the Thesis  5
CHAPTER 2  RELATIONSHIP AND MOTION CONTROL OF PTZ CAMERAS…………………………………………………….…..7
2.1    Projective Geometry  7
2.2    The Direct Linear Transformation (DLT) Algorithm  9
2.3    Pan/tilt control method of the dynamic PTZ camera  13
CHAPTER 3  TEMPLATE MATCHING IN THE DYNAMIC PTZ        VIEW…………………………………………………………………...21
3.1    Template Model Generation  21
3.2    Template Matching  23
CHAPTER 4  COOPERATIVE TRACKING SYSTEM IN MULTIPLE VIEWS……………………...……………………………29
4.1    Tracking Human and Car Objects in a Scene  29
4.2    Two Tracking Algorithms of the Dynamic Camera  29
CHAPTER 5  EXPERIMENT RESULTS……………..……………34
5.1    Experimental Results of the human tracking  35
5.2    Experimental Results of the car tracking  38
CHAPTER 6  CONCLUSIONS AND FEATURE WORKS….……39
REFERENCES………………………………………………………...40

LIST OF FIGURES
Fig. 1.1  The system flow diagram…………………………………………………...4
Fig. 2.1  The projective transformation between two projective planes…….……….8
Fig. 2.2  The corresponding pairs between two camera views..……………………12
Fig. 2.3  Top view of the simplified camera model…………………………….......14
Fig. 2.4  Trajectories of the different points while the dynamic camera is pure panning…………………………………………………………………………….....17
Fig. 2.5  Relationships between the points after panning step...................................18
Fig. 2.6  Coordinate variation of the pixel during the dynamic camera tilting and panning steps…………………………………………………………........…………19
Fig. 3.1  The static and dynamic camera view during the target detection process………………………………………………….....………………………….23
Fig. 3.2  Shift value between the matching blocks…………………………………24
Fig. 3.3  The color matching order of the block at different position………………26
Fig. 3.4  Three different matching template sizes at one block position for color matching……………………………………………………………….……………..28
Fig. 4.1  Initial views of the zero positions of (a) static camera and (b) dynamic camera……………………………………………………………………………..…30
Fig. 4.2  (a) dynamic camera view at t=31sec. (b) dynamic camera view at t=33sec. (c) motion difference between 4.2 (a) and 4.2 (b)……………………………...…….32
Fig. 4.3  The related parameters for determining the moving speed and moving direction of the tracking object……………………...………………………………..33
Fig. 5.1  Testing environment………………………………………………………34
Fig. 5.2  Experiment result of the human tracking (I)…………………………...….36
Fig. 5.3  Experiment result of the human tracking (II)………………………….….37
Fig. 5.4  Experiment result of the human tracking (III)…………………………….38
Fig. 5.5  Experiment result of the car tracking……………………………….……..38

                                

References
[1] Ser-Nam Lim, Ahmed Elgammal and Larry S. Davis, “Image-Based Pan-Tilt Camera Control in a Multi-Camera Surveillance Environment.” Multimedia and Expo, 2003. ICME '03. Proceedings. 2003 International Conference on, Volume 1, July 2003
[2] Y. Wu, T. Yu, and G. Hua, “Tracking Appearances With Occlusions”, Proceedings IEEE CVPR 2003, Volume 1, June 2003, pp. 789-795.
[3] M. Hu, W. Hu, and T. Tan, “Tracking People Through Occlusions”, ICPR Aug. 2004, pp.724 – 727.
[4] Ser-Nam Lim, Larry S. Davis and Ahmed Elgammal, “Scalable Image-Based Multi-Camera Visual Surveillance System”, Proceedings. IEEE Conference on Advanced Video and Signal Based Surveillance, July 2003, pp.205 – 211.
[5] A.W. Senior, A. Hampapur, M. Lu, “Acquiring Multi-Scale Images by Pan-Tilt-Zoom Control and Automatic Multi-Camera Calibration”, Seventh IEEE Workshops on Application of Computer Vision (WACV/MOTION'05), Volume 1, pp. 433-438
[6] Sangkyu Kang, Joonki Paik, Andreas Koschan, Besma Abidi, and Mongi A. Abidi, “Real-time video tracking using PTZ cameras” Proc. of SPIE 6th International Conference on Quality Control by Artificial Vision, Vol. 5132, pp. 103-111, Gatlinburg, TN, May 2003.
[7] Christopher R. Wren, U. Murat Erdem, Ali J. Azarbayejani, “Automatic Pan-Tilt-Zoom Calibration in the Presence of Hybrid Sensor Networks”, International Multimedia Conference Proceedings of the third ACM international workshop on Video surveillance & sensor networks, 2005, pp.113 – 120.
[8] Gin-Shu Young, Tsai-Hong Hong, Martin Herman, and Jackson C. S. Yang, “Kinematic Calibration of an Active Camera System”, CVPR Jun. 1992, pp.748 – 751.
[9] Lourdes de Agapito, Richard I. Hartley, Eric Hayman “Linear Self-Calibration of a Rotating and Zooming Camera”, CVPR’99, Vol. 1, p1015.
[10] Takashi Matsuyama, Norimichi Ukita “Real-Time Multi-Target Tracking by a Cooperative Distributed Vision System”, International Conference on Autonomous Agents Proceedings of the first international joint conference on Autonomous agents and multiagent systems: part 2, 2002, pp.829 – 838
[11] L. De Agapito, E. Hayman and I. Reid, “Self-calibration of rotating and zooming cameras”, International Journal of Computer Vision 45, pp. 107-127. 2001
[12] Sudipta N. Sinha and Marc Pollefeys, “Towards Calibrating a Pan-Tilt-Zoom Camera Network”, OMNIVIS 2004, workshop on Omnidirectional Vision and Camera Networks held in conjunction with ECCV 2004
[13] I. Everts, N. Sebe, G.A. Jones, “Cooperative Object Tracking with Multiple PTZ Cameras”, International Conference on Image Analysis and Processing, Sep. 2007.
[14] Mircea Nicolescu and Gérard Medioni, “Electronic Pan-Tilt-Zoom: A Solution for Intelligent Room System”, Proc. ICME'2000, pp. 1581-1584
[15] P. Pérez, C. Hue, J. Vermaak, and M. Gangnet, “Color-Based Probabilistic Tracking”, In European Conference on Computer Vision, 2002, number 2350 in Lecture Notes in Computer Science, pp. 661–675.
[16] K. Nummiaro, E. Koller-Meier, and L.Van Gool, “An Adaptive Color-Based Particle Filter”, IVC, Vol. 2(1), 2003, pp.99-110.

全文公開日期本全文未授權公開 (校內網路)
全文公開日期本全文未授權公開 (校外網路)

簡易檢索 / 詳目顯示

相關論文