臺灣博碩士論文加值系統

在本篇論文當中，我們提出了一個新演算法可以自動從視訊切割人物並替換影像背景得到一組新的影像序列。在這個任務中，如何快速且有效的取得人物分割結果是最主要的挑戰。首先，我們提出在視訊中追蹤人物動作的訊息來改進Random Walk演算法中的先前形狀模型，利用時間上的一致性來保留人型的完整。另外，我們合併亮度與邊緣資訊差異於節點間的權重定義並取能量最小化讓相似的點盡量收斂到同樣的分割。我們的實驗結果展現出我們可以有效率的獲得相當準確的人物分割結果。
另外，我們在多核心平台PACDuo上實作系統的初步架構，基於平台資源有限，我們提出使用TYPE和INDEX的演算法來有效降低計算量，最後再使用針對多核心的資源配置策略與局部的資料傳輸到DSP核心與ARM處理器同步運算。最後的實驗結果顯示兩種方法都有效的降低系統在多核心處理平台執行時間。

In this thesis, we propose an automatic video conferencing system for background substitution. Since humans are the principal subject in these videos, our framework is based on human shape clues to separate humans from complex background and replace or blur the background for immersive communication. We first detect face position and size, track human boundary across frames, and propagate the segmentation likeihood to the next frame for obtaining the trimap to be used as input to the Random Walk algorithm. Besides, we also include gradient magnitude in edge weight to enhance the Random Walk segmentation results. In this part, we demonstrate the effectiveness of the proposed background substitution system through experiments on some real videos.
We also present a system based on a multi-core processing architecture. Two tables, TYPE and INDEX, are introduced to fast locate the required data for the close-form solution. We demonstrate the parallelization strategies for the proposed fast RW algorithm and face detection on heterogeneous multi-core embedded platform to make the most use of the system architecture. Compared to the single processor implementation, the experimental results show significant speedup of the parallelized human background substitution system on a multi-core embedded platform, which consists of an ARM processor and two DSP cores.

Abstract iv
Contents i
List of Figures iii
List of Tables iv
Chapter 1. Introduction 1
1.1. Problem Description and Motivation 1
1.2. System Overview 3
1.3. Main Contribution 4
1.4. Thesis Outline 5
Chapter 2. Related Work 6
2.1. Approaches have Learned Background Model 6
2.2. Approaches under Stationary Camera 9
2.3. Approaches with Human Shape Template 9
Chapter 3. Proposed Method 13
3.1. Human Shape Prior Adaption in a Single Image 13
3.2. Human Shape Prior Adaption for Video 15
3.2.1. Boundary Estimating 17
3.2.2. Gray Area Determination 18
3.2.3. Trimap Construction for Random Walk Segmentation 19
3.3. Foreground Segmentation 21
3.3.1. Generative Graphic Model 22
3.3.2. Edge Weight 22
3.3.3. Likelihood Estimation 23
3.3.4. Convex optimization 24
3.4. Applications related to Human Segmentation 25
3.4.1. Background substitution 26
3.4.2. Background blurring 27
Chapter 4. Parallelized System on Multi-core Embedded Platform 29
4.1. Overview of the Platform 29
4.2. Fast Random Walk by tables 31
4.2.1. Gauss-Seidel Iteration 31
4.2.2. Data Arrangement 32
4.2.3. TYPE and INDEX Tables 33
4.3. Parallelization Strategies 35
4.3.1 Data Allocation 35
4.3.2. Data Transmission 36
Chapter 5. Experimental Results 38
5.1. Performance on PC version 38
5.2. Performance on PACDuo 46
Chapter 6. Conclusion 48
Bibliography 49

[1] C. Stauffer and W.E.L Grimson, “Adaptive background mixture models for real-time tracking,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), 1999
[2] J. Sun, W. Zhang, X. Tang, and H.Y. Shum, “Background Cut,” In Proceedings of European Conference on Computer Vision (ECCV), 2006, pp. 628–641.
[3] S. Kwak, I. Park, J. Lee, H. Byun, and G. Bae, “Automatic Background Substitution using Monocular Camera and Temporal Foreground Probability Model,” Proceeding of 2nd International Conference on Ubiquitous Information Management and Communication (ICUIMC), 2008, pp. 506-510
[4] F. J., H. Lopez, and M. Rivera, “Binary Segmentation of Video Sequences in Real Time,” Ninth Mexican International Conference on Artificial Intelligence (MICAI), 2010, pp. 163-168,
[5] A. Criminisi, G. Cross, A. Blake, and V. Kolmogorov, “Bilayer Segmentation of Live Video,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), 2006, pp. 53-60.
[6] P. Yin, A. Criminisi, J. Winn, and I. Essa, “Bilayer Segmentation of Webcam Videos Using Tree-Based Classifiers,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, vol. 33, pp. 30-42.
[7] F. Zhong, X. Qin, and Q. Peng, “Transductive Segmentation of Live Video with Non-stationary Background,” IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2010, pp. 2189-2196.
[8] C. Zhang, Y. Rui, and L.W. He, “Light Weight Background Blurring for Video Conferencing Applications,” IEEE International Conference on Image Processing (ICIP), 2006, pp. 481-484.
[9] B. Ding, R. Shi, Z. Liu, and Z. Zhang, “Human object segmentation using Gaussian mixture model and graph cuts,” International Conference on Audio Language and Image Processing (ICALIP), 2010, pp. 787-790.
[10] A. Parolin, G.P. Fickel, C.R. Jung, T. Malzbender, and R. Samadani, “BILAYER VIDEO SEGMENTATION FOR VIDEOCONFERENCING APPLICATIONS,” IEEE International Conference on Multimedia and Expo (ICME), 2011, pp. 1-6.
[11] K.C. Li, and S.H. Lai. “Automatic pedestrian image segmentation by using human shape prior,”.
[12] http://research.microsoft.com/vision/cambridge/i2i/DSWeb.htm.
[13] L. Grady, “Random walks for image segmentation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, vol. 28(11), pp. 1768–1783.
[14] R. Merris, “Laplacian Matrices Of Graphs: A Survey,” Linear Algebra and its Applications, 1994, vol. 197-198, pp. 143–176.
[15] D.C.W. Chang, I.T. Liao, J. K. Lee, W.F. Chen, S.Y. Tseng, and C.W. Jen, “PAC DSP CORE AND APPLICATION PROCESSORS,” IEEE International Conference on Multimedia and Expo, 2006, pp. 289 –292.
[16] R. Barrett, M. Berry, T.F. Chan, J. Demmel, J.M. Donato, J. Dongarra, V. Eijkhout, R. Pozo, C. Romine, and H.V.D. Vorst, “Templates for the Solution of Linear Systems,” Philadelphia, PA: SIAM, 1994, ed. 2nd