跳到主要內容

臺灣博碩士論文加值系統

(44.200.140.218) 您好!臺灣時間:2024/07/18 03:57
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:林哲緯
研究生(外文):Chewei Lin
論文名稱:無標記式人體動作擷取技術
論文名稱(外文):無標記式人體動作擷取技術
指導教授:盧天麒盧天麒引用關係
指導教授(外文):Tainchi Lu
學位類別:碩士
校院名稱:國立嘉義大學
系所名稱:資訊工程學系研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
畢業學年度:104
語文別:中文
論文頁數:54
中文關鍵詞:角色動畫無標記式動作擷取動作追蹤姿勢預測深度學習
外文關鍵詞:Character AnimationMarkerless Motion CaptureSum of GaussiansMotion PredictionDeep learning
相關次數:
  • 被引用被引用:0
  • 點閱點閱:374
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
近年來,遊戲與動畫產業蓬勃的發展,對於虛擬角色皮膚紋理的呈現及運動的感受都希望越來越擬真,傳統的做法由專業的演員穿戴動作擷取器(motion capture, MoCap)作出腳本的動作,但是整套的設備非常昂貴且演員必須穿戴道具服較不方便,因此無標記式的動作擷取技術的開發,可提供一般使用者以低成本的價格就能建置屬於自己的腳色動畫,幫助使用者降低動畫與遊戲開發上的成本。
本研究提出了通過使用低成本的靜態攝影機進行無標記式人體動作捕捉。我們利用鉸鏈式人體骨架代表複雜的人體,並採用了三個維度的攝影機拍攝方法取得完整的影像資訊且不需深度感測器的輔助來取得深度資訊。為了達到即時性,本研究採用顏色作為辨識的特徵,使用高斯建立影像模型並使用深度學習進行特徵分析,從建立的資料庫中尋找可能的人體姿勢,接著透過鉸鏈式骨架的輔助進行動作預測,進一步達到動作即時追蹤的效果,最後採用標準的BVH格式來存儲連續的運動數據,並提供其他系統進一步的使用。
本論文所開發的動作擷取系統,可以準確的將使用者的動作擷取出來,並解決Kinect 無法自身旋轉的問題,即便使用者不是專業的動畫師或使用者沒有能力購置昂貴的光學式動作擷取器,也可以使用本研究所開發的系統製作屬於自己的三維角色動畫與專屬的互動體驗。本研究未來將可以運用在特殊的技能訓練或是互動教學等相關應用上,使人們參與模擬體驗能更加的普及。
Depending on the rapid development of the game and animation industry in recent years, the character’s skin texture rendering and human movement are hoped more realistic. Traditional motion capture can capture the real motion by actor, but this system is expensive and not comfortable for actor to wear. Therefore, markerless humanoid motion capture technology is developed. This technology can creates own character with lower price and helps user reduce costs for the development of game and animation.

The paper presents an easy-to-implement technique of performing a markerless humanoid motion capture by using a low-cost and stationary camera. To achieve the efficacy of real time calculation, we use articulated simplified skeleton to represent a complicated human body and do not require any auxiliary multiple depth sensors for obtaining depth information. This system uses color information as characteristic of recognition, we use Gaussians to set up the image model and use deep learning algorithm to match key pose. Next, according to the articulated skeleton to help the motion prediction and compare with the range of the projection and image model of human’s pose. Finally, the continuously motion data will be saved by standard BVH format, offering other systems to use in the future.

The motion capture system in this paper not only can captures user’s motion accurately but also solves the problem of Kinect that can’t rotate by itself. Even though the user is not a professional animator or unable to buy expensive motion capture system, everyone can have his own 3D character animation by this markerless humanoid motion capture. This research can be used on special skills training or interactive teaching and even other applications in the future. It will let people participate the simulated experience commonly.
第一章 緒論 - 6 -
1.1 動機與目的 - 6 -
1.2 範圍與構想 - 7 -
1.3 貢獻與限制 - 8 -
第二章 相關研究 - 10 -
2.1 動作擷取技術 - 10 -
2.2 人體骨架模型 - 12 -
2.3 深度學習應用於動作資料 - 13 -
第三章 系統架構與實作 - 15 -
3.1攝影機校正 - 15 -
3.2 建立人體骨架資料 - 18 -
3.3 相似性計算 - 19 -
3.4人體模型建立 - 21 -
3.5 人體姿勢特徵抽取 - 23 -
3.6 深度學習分析特徵 - 24 -
3.7 人體姿勢估計 - 26 -
3.8動作追蹤 - 27 -
第四章 實驗結果 - 30 -
第五章 結論與未來展望 - 47 -
參考文獻 - 48 -
[1] L. Ballan, and G. M. Cortelazz, “Marker-less Motion Capture of Skinned Models in a Four Camera Set-up using Optical Flow and Silhouettes” In 3D Data Processing, Visualization and Transmission, Vol 27, 2008.

[2] A. Bissacco, M.-H. Yang, and S. Soatto, “Fast human pose estimation using appearance and motion via multi-dimensional boosting regression,” In Computer Vision and Pattern Recognition, pp. 1-8, 2007.

[3] T. Brox, B. Rosenhahn, J. Gall, and D. Cremers, “Combined region and motion-based 3d tracking of rigid and articulated objects,” Transactions on Pattern Analysis and Machine Intelligence, Vol. 32, pp. 402–415, 2010.

[4] L. Bo, C. Sminchisescu, “Twin gaussian processes for structured prediction,” International Journal of Computer Vision, Vol. 87, Issue 1, pp. 28-52, 2010.

[5] M. Bray, P. Kohli, and P. Torr, “Posecut: Simultaneous segmentation and 3d pose estimation of humans using dynamic graph-cuts,” In European Conference on Computer Vision, pp. 642–655, 2006.

[6] S. Choppin, Engineering sport: the Centre for Sports Engineering Research. available through http://engineeringsport.co.uk/2011/05/09/kinect-biomechanics-part-1/ , Dec. 2015.

[7] F. Caillette, A. Galata, and T. Howard, “Real-time 3-D human body tracking using learnt models of behaviour.” Comput Vis Image Underst. Vol. 109, No. 2 pp.112–125, 2008

[8] Y. Du, W. Wang, and L. Wang, “Hierarchical recurrent neural network for skeleton based action recognition,” In Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2015.

[9] H. Erives, and G.J. Fitzgerald, “Automatic subpixel registration for atunable hyperspectral imaging system.” IEEE Geoscience and Remote Sensing Letters, Vol. 3, No. 3, pp. 397-400, 2006.

[10] W. Förstner, and E. Gülch. “A fast operator for detection and precise location of distinct points, corners and centres of circular features.” In ISPRS Intercommission Workshop, interlaken, 1987.

[11] J. Gall, B. Rosenhahn, and H.-P. Seidel, “Drift-free tracking of rigid and articulated objects,” In Computer Vision and Pattern Recognition, 2008.

[12] A. Goshtasby, “Image registration by local approximation methods.” Image and Vision Computing, Vol. 6, No. 4, pp. 255-261, 1988

[13] J. Gall, C. Stoll, E. De Aguiar, H.P. Seidel, C. Theobalt, and B. Rosenhahn, “Motion capture using joint skeleton tracking and surface estimation.” IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops, pp.1746–1753, 2009.

[14] Graves, A., Mohamed, A.-R., and Hinot, G. “Speech recognition with deep recurrent neural networks,” In Proc. of Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on, IEEE, pp. 6645–6649, 2013.

[15] N. Hasler, B. Rosenhahn, T. Thorm¨ahlen, M. Wand, J. Gall, and H.-P. Seidel, “Markerless motion capture with unsynchronized moving cameras,” In Computer Vision and Pattern Recognition, 2009

[16] N. Hasler, B. Rosenhahn, T. Thorm¨ahlen, M. Wand, J. Gall, and H.-P. Seidel, “ Markerless motion capture with unsynchronized moving cameras,” In Computer Vision and Pattern Recognition, 2009.

[17] D. Holden, J. Saito, T. Komura, and T. Joyce, “Learning motion manifolds with convolutional autoencoders,” In SIGGRAPH Asia 2015 Technical Briefs, ACM, Vol. 18, No. 4, pp. 1–18, 2015.

[18] Harris and M. Stephens. “A combined corner and edge detector”. Proceedings of the 4th Alvey Vision Conference, pp. 147--151, 1988.

[19] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks.” In Proc. of Advances in Neural Information Processing Systems, pp. 1097–1105, 2012.

[20] P. Kohli, J. Rihan, M. Bray, and P.H.S. Torr, “Simultaneous segmentation and pose estimation of humans using dynamic graph cuts.” Int J Comput , Vol. 79, No3 pp. 285–298, 2008.

[21] C. S. Lee, and A. Elgammal, “Coupled visual and kinematic manifold models for tracking,” International Journal of Computer Vision, Vol. 87, Issue 1, pp. 118-139, 2010.

[22] R. Li, T. P. Tian, S. Sclaroff, and M. H. Yang, “3d human motion tracking with a coordinated mixture of factor analyzers,” International Journal of Computer Vision, Vol. 87, Issue 1, pp. 170-190, 2010.

[23] D. Moschini,A. Fusiello,. “Tracking human motion with multiple cameras using an articulated model.” Computer vision/computer graphics collaboration techniques, proceedings, 1–12, 2009.

[24] K. Ogawara, X. Li, and K. Ikeuchi, “Marker-less human motion estimation using articulated deformable model.” Proceedings of the IEEE International Conference on Robotics and Automation, Vol. 1 No. 10, pp. 46–51, 2007.

[25] P. Ronald, “Vision-based human motion analysis: An overview,” Journal of Computer Vision and Image Understanding, Vol. 108, Issue 1, pp. 4-18, 2007.

[26] C. Schmaltz, B. Rosenhahn, T. Brox, and J. Weickert, “Region-based pose tracking with occlusions using 3d models,” Machine Vision and Applications, pp. 1–21, 2011.

[27] L. Sigal, A. Balan, and M. Black, “HumanEva: Synchronized Video and Motion Capture Dataset and Baseline Algorithm for Evaluation of Articulated Human Motion,” International Journal of Computer Vision, Vol. 87, Issue 1, pp. 4-27, 2010.

[28] C. Stoll, N. Hasler, J, Gall, H.-P. Seidel, and C. Theobalt, “Fast Articulated Motion Tracking using a Sums of Gaussians Body Model,” In International Conference on Computer Vision, pp. 951-958, 2011.

[29] A. Sundaresan, R. Chellappa, “Model driven segmentation of articulating humans in Laplacian Eigenspace. Pattern Anal Mach” Intell IEEE Trans, Vol. 30, No. 10 pp. 1771–1785, 2008.

[30] L. Taycher, D. Demirdjian, T. Darrell, and G. Shakhnarovich, “Conditional random people: Tracking humans with crfs and grid filters,” In Computer Vision and Pattern Recognition, Vol. 1, pp. 222-229, 2006.

[31] C. Theobalt, I. Albrecht, J. Haber, M. Magnor, and H.-P. Seidel, “Pitching a baseball: tracking high-speed motion with multi-exposure images,”ACM Transactions on Graphics, Vol. 23, Issue 3, pp. 540-547, 2004.

[32] G. W. Taylor, and G. E. Hinton, “Factored conditional restricted boltzmann machines for modeling motion style,” In Proc. of the 26th International Conference on Machine Learning, ACM, pp. 1025–1032, 2009.

[33] M. Thomas, A. Hilton, and V. Kr¨uger. “A survey of advances in vision-based human motion capture and analysis,” Computer Vision and Image Understanding, Vol. 104, Issue 2, pp. 90–126, 2006.

[34] M. Van den Bergh, E. Koller-Meier, and L. Van Gool, “Realtime body pose recognition using 2d or 3d haarlets,” International Journal of Computer Vision, Vol. 83, Issue 1, pp. 72–84, 2009.

[35] Z. Xiaoyun, L. Yuncai, L., and T.S. Huang, Motion analysis of articulated objects from monocular images. Pattern Anal Mach Intell IEEE Trans. Vol. 28, No. 4, pp. 625–636, 2006.

[36] S. Yonemoto, D. Arita, and R. Taniguchi, “Real-time human motion analysis and ik-based human figure control,” In Workshop on Human Motion, pp. 149–154, 2000.

[37] C. Wren, A. Azarbayejani, T. Darrell, and A. Pentland, “Pfinder: Real-time tracking of the human body. TPAMI,” Vol. 19, pp. 780–785, 1997.

[38] B. Zitova´, J. Flusser, “Image registration methods: a survey,” Institute of Information Theory and Automation, Academy of Sciences of the Czech Republic Vol. 4 pp. 137-158 2003.

[39] Sylvia X.M. Yang, Martin S. Christiansen, Peter K. Larsen, Tine Alkjær, Thomas B. Moeslund, Erik B. Simonsenc and Niels Lynnerup, “Markerless motion capture systems for tracking of persons in forensic biomechanics:an overview.” Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, Vol. 2, No. 1, pp. 46–65, 2014.
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top