臺灣博碩士論文加值系統

隨著科技的發展，數位學習已是越來越多人接受並採用的學習方式，但現今的數位學習多是靜態知識類的課程，針對肢體動作的教學系統則較不普及。因此，本論文提出一個針對肢體動作的互動教學之方法，透過影像追蹤，記錄人體在立體空間中的移動軌跡，讓使用者能依據軌跡做出相同的動作來達到學習的效果。本系統主要流程有：影像前置處理、人體辨識、人體部位追蹤、軌跡簡化、軌跡的校正、動作判定等。
本論文將研究如何正確的分辨人體的各個主要動作的關鍵部位，利用這些關鍵部位的位置即可建構出人體主要的骨架，而記錄這些部位的軌跡即可相當於記錄一段動作。此外，紀錄於立體空間的軌跡會依據使用者的肢體長短大小、與感測器距離或角度不同做出對應的調整，達到更正確的學習效果，此部分則會利用到前面得到的骨架長短、角度再透過數學運算得出轉換後的軌跡。在成果中，使用者可以看見每一段動作的軌跡顯示於教學畫面中並跟著做，系統會判斷肢體部位移動的軌跡是否有在標準軌跡的範圍內來評分，並提出需要改善的部分或是重複錯誤特別多的片段，讓學習者可以反覆學習修正。
本研究主要以影像跟深度感測做運算處理，使用者無需額外穿戴其他設備，可以不受干擾的做動作，成本也較其他動作捕捉的技術（如：標籤、穿戴式感測器等方式）便宜，對於這種數位教學模式的推廣有正面且實質的助益。而研究中的動作捕捉與動作比對的技術，在未來應用於人機互動上也是相當方便直覺的方式。

In the past, for the domain of human movement, sports or dance learners relied heavily on experts to judge whether or not learners’ body movements were correct, and offered suggestions for improvement. Today, depends on science and technology, analyses via depth image sensor can detect the motion of each body region. According to the image sensor tracking result, system can measure the efforts made by learners’ movements. This paper presents a novel system provides an augmented reality interactive body movements training program. Through the image tracking algorithm, the path of movement of the human body in the three-dimensional space are recorded which act as a standard movement of a virtual coach. Learners can follow the pre-recorded coach's actions from display screen to imitate the correct body movements step by step. Learner can make the same motion according to the body movements of virtual coach to achieve the effect of learning. It also provides a learning mechanism that can correctly distinguish the key parts of the main motion of the human body, and can repeat the training many times for correcting the wrong body movement of learners. This kind of movement learning, in which learner acquire knowledge by doing and via reflection on the virtual coach's actions, is full of movement, imagination, and self-directed play. Even without the advice of a real expert or coach, the system can provide advice and training program for body movement learners. Through this system, the learner can use the way of body movement self-learning at any place.

摘要 I
Abstract II
致謝 III
目錄 IV
圖目錄 V
第一章. 緒論 1
1.1. 研究動機 1
1.2. 研究目的 2
1.3. 論文大綱 2
第二章. 文獻探討 3
2.1. 動作捕捉與人體追蹤 3
2.2. Kinect for windows感測器 5
2.3. 格拉斯-普克演算法（Douglas–Peucker algorithm） 6
2.4. 積分影像(integral image) 7
2.5. Haar-like feature之人臉檢測法 8
第三章. 研究方法 10
3.1. 流程架構 11
3.2. 影像去背 13
3.3. 人體辨識與追蹤 16
3.4. 簡化軌跡 17
3.5. 軌跡校正 19
3.6. 生成引導軌跡 22
3.7. 動作比對設計 24
3.8. 互動情境 26
第四章. 實驗結果 27
4.1. 實驗環境 27
4.2. 實驗結果 28
第五章. 結論與未來展望 34
參考文獻 35

[1]
Jamie Shotton, Andrew Fitzgibbon, Mat Cook, Toby Sharp, Mark Finocchio, Richard Moore, Alex Kipman, Andrew Blake. “Real-Time Human Pose Recognition in Parts from a Single Depth Image.” CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition, Pages 1297-1304.

[2]
Douglas, D. and Peucker, T. “Algorithms for the Reduction of the Number of Points Required to Represent a Digitized Line or its Caricature.” The Canadian Cartographer, Vol. 10, No. 2, Pages 112-122, 1973.

[3]
Gang Hu and Qigang Gao. “3D perceptual shape feature-based body parts classification and pose estimation.” J-HGBU '11 Proceedings of the 2011 joint ACM workshop on Human gesture and behavior understanding, Pages 75-76, 2011.

[4]
John Hershberger and Jack Snoeyink. “An O(nlogn) implementation of the Douglas-Peucker algorithm for line simplification.” Proceedings of the tenth annual symposium on Computational geometry, Pages 383-384, 1994.

[5]
Takeshi Mita, Toshimitsu Kaneko, and Osamu Hori. “Joint Haar-like Features for Face Detection.” ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision, Vol 2, Pages 1619 -1626, 2005.

[6]
Phillip Ian Wilson and John Fernandez. “Facial feature detection using Haar classifiers.” Journal of Computing Sciences in Colleges, Vol 21, Issue 4, Pages 127-133, 2006.

[7]
Alper Yilmaz, Omar Javed, and Mubarak Shah. “Object tracking: A survey. ” ACM Computing Surveys (CSUR), Vol. 38, Issue 4, 2006.

[8]
Alan J. Lipton, Hironobu Fujiyoshi, and Raju S. Patil. “Moving Target Classification and Tracking from Real-time Video.” The 4th IEEE Workshop on Applications of Computer Vision (WACV'98), Page 8, 1998.

[9]
Ming-I Chien. “Posture Interactive Learning System Based on Human Action Tracking.” 2016.

[10]
Shaikh, Soharab Hossain, Saeed, Khalid,Chaki, Nabendu. “Moving Object Detection Approaches, Challenges and Object Tracking.” Moving Object Detection Using Background Subtraction, Pages 5-14, 2014.