臺灣博碩士論文加值系統

在本篇論文中，我們提出一套應用於多台攝影機之多角度人臉偵
測系統。此系統可根據多攝影機擷取的影像偵測出影像中的人臉位
置，並得到在三維空間中人臉方向的鳥瞰圖。有別於以往的作法，我
們並不在二維的影像中直接作搜尋與偵測、或是將這些在二維的影像
中的偵測結果投影到三維空間中作結合；反之，我們的系統直接在三
維空間中進行搜尋，並將三維空間投影回二維影像中加以比對處理，
以判斷在這個三維空間中是否存在著某種方向的人臉。這樣的做法使
得我們的系統可以有效地結合多攝影機中的二維影像資訊，並可避免
以往因在二維影像中的錯誤偵測所導致的不明確資訊整合。

In this paper, we propose a multi-view face detection system, which
is capable of detecting all targets’ faces in the given images and is able to
illustrate the bird-eye view direction of each face in the 3-D space in a
multi-camera surveillance system. Unlike existing approaches, the
proposed system does not directly detect targets over the 2-D image
domain nor project the 2-D detection results back to the 3-D space for
correspondence. Instead, our system searches for the targets over small
cubes in the 3-D space. Each searched 3-D cube is projected onto the 2-D
camera views to determine the existence and direction of human faces.
This approach can help us to efficiently combine 2-D information from
different camera views and to suppress the ambiguity caused by 2-D
detection errors.

Chapter 1. Introduction ........................................................................................ 1
Chapter 2. Backgrounds ...................................................................................... 3
2.1. Face Detection ................................................................................... 3
2.2. Multi-View Face Detection................................................................. 5
2.3. Multi-Camera Face Detection ............................................................. 6
Chapter 3. Proposed Method ............................................................................... 9
3.1. 2-D Detection to 3-D Detection ........................................................ 10
3.1.1. 2-D Face Detection Framework ................................................ 10
3.1.2. 3-D Face Detection Framework ................................................. 11
3.2. 3-D Position Estimation ................................................................... 13
3.2.1. Background Subtraction ........................................................... 14
3.2.2. Information Fusion ................................................................... 15
3.3. Environment Setup ........................................................................... 18
3.3.1. Orientation of the Scenario ....................................................... 18
3.3.2. Assumption and Outcome ......................................................... 19
3.3.3. Hypothesis Definition ............................................................... 21
3.4. Multi-View Face Detection............................................................... 23
3.4.1. Viola and Jones’ Algorithm ....................................................... 24
3.4.2. Modified Face Detection Algorithm .......................................... 25
3.4.3. Eight Views Detectors .............................................................. 26
3.5. Information Fusion And Result Inference ......................................... 27
3.6. Overall System ................................................................................. 32
Chapter 4. Experimental Results ........................................................................ 33
4.1. Multi-view Face Dataset ................................................................... 33
4.2. Experimental Results ........................................................................ 36
Chapter 5. Conclusions ...................................................................................... 42
Chapter 6. References........................................................................................ 43

[1] Ching-Chun Huang and Sheng-Jyh Wang, “Moving Targets Labeling and
Correspondence over Multi-Camera Surveillance System Based on Markov
Network.” IEEE International Conference on Multimedia and Expo, June
28-July 3, 2009.
[2] R.E. Schapire, Y. Freund, P. Bartlett, and W.S. Lee, “Boosting the Margin: A
New Explanation for the Effectiveness of Voting Methods.” Proc. Fourth
International Conference Machine Learning, p322-330, 1997.
[3] P. Viola and M. Jones, “Rapid Object Detection Using a Boosted Cascade of
Simple Features.” Proc. IEEE Conference on Computer Vision and Pattern
Recognition, p511-p518, 2001.
[4] M. Jones and P. Viola, “Fast Multi-view Face Detection.” IEEE Conference on
Computer Vision and Pattern Recognition, pp. 2, TR2003-96, July 2003.
[5] P. Viola and M. Jones, “Robust Real-Time Face Detection.” International Journal
of Computer Vision 57(2), p137-p154, 2004.
[6] B. Wu, H. Ai, C. Huang, and S. Lao, “Fast rotation invariant multi-view face
detection based on real adaboost.” IEEE Conference on Automatic Face and
Gesture Recognition, 2004.
[7] C. Huang, H. Ai, Y. Li, and S. Lao, “Vector Boosting for Rotation Invariant
Multi-View Face Detection.” IEEE International Conference on Computer
Vision , pp.446-453, Beijing, China, Oct 17-20, 2005
[8] C. Huang, H. Ai, Y. Li, and S. Lao, “High-performance rotation invariant
multiview face detection.” IEEE Transactions on Pattern Analysis and Machine
Intelligence, vol.29, no.4, pp.671–686, 2007.
[9] Z. Zhang, G. Potamianos, M. Liu, and T.S. Huang, “Robust multi-view
multicamera face detection inside smart rooms using spatio-temporal dynamic
programming.” IEEE Conference on Automatic Face and Gesture Recognition,
2006.
[10] Z. Zhang, G. Potamianos, A.W. Senior, and T.S. Huang, “Joint face and head
tracking inside multi-camera smart rooms.” Signal, Image and Video Processing,
vol. 1, pp. 163–178, 2007.
[11] Saad M Khan and Mubarak Shah, “Tracking Multiple Occluding People by Localizing on Multiple Scene Planes.” IEEE Transactions on Pattern Analysis
and Machine Intelligence, Vol. 31, pp. 505-519, March 2009.
[12] S. Li, Z. Zhang, L. Zhu, H.-Y. Shum, and H. Zhang, “Floatboost learning for
classification.” In Proceedings of The 16-th Annual Conference on Neural
Information Processing Systems. Vancouver, Canada, December 9-14, 2002.
[13] Y. Kobayashi, D. Sugimura, and Y. Sato, “3D head tracking using the particle
filter with cascaded classifiers.” In Proc. of British Machine Vision Conference,
pages 37–46, September 2006.
[14] Ching-Chun Huang and Sheng-Jyh Wang, “A Monte Carlo Based Framework for
Multi-Target Detection and Tracking Over Multi-Camera Surveillance System.”
European Conference on Computer Vision Workshop on Multi-camera and
Multi-modal Sensor Fusion Algorithms and Applications, Marseille, France,
October 12-18, 2008.
[15] Ching-Chun Huang and Sheng-Jyh Wang, “A Bayeisan Hierarchical Framework
for Multi-Target Labeling and Correspondence with Ghost Suppression over
Multi-Camera Surveillance System.” submitted to IEEE Transactions on
Automation Science and Engineering. (Revision)