本研究期望建立一套規則式邏輯作為手勢辨識之基礎，以最有效率的方式達成手勢辨識的目的。研究方法是透過一些自然手勢的表達來找出距離及角度的相關特徵值，選定的手勢是由中文手語(Chinese Sign Language)中挑選20個手勢。受試者共有10人，分為訓練組(5人)及測試組(5人)，由訓練組建立手勢資料庫，再藉由測試組的資料進行驗證。實驗設計是藉由VICON攝影機抓取受試者手指關節上感光點在空間的3D座標位置，將此資料經過特徵值擷取並正規化後，藉由觀察法找出手勢分類的重要特徵，並用來作為規則式邏輯建構的基礎，以發展手勢辨識系統(Finger Gesture Recognition System, FGRS)，進而完成手勢之辨識。
由實驗之結果說明最佳分類手勢之特徵值個數僅需要7個即能達到100%分類，分別為中指、無名指、小指的絕對距離(NADD3、NADD4、NADD5)和大拇指-食指、大拇指-中指的相對距離(NRD1、NRD2)和大拇指、食指的近端指間關節彎曲角度(JAP1、JAP2)。手勢被辨識的最佳區間為”平均值±2個標準差”，在【p-大拇指＆食指相對位置】、【p-食指伸直】、【p-小指伸直】三類規則式邏輯之最佳辨識能力為70%，而新規則式邏輯之辨識能力為58%，雖然新規則式邏輯之辨識率不如其他三類，但由於其不須經過細微之敏感度調整，不僅增加效率，也證明本研究歸納出之通用性準則有其存在價值。此外，由誤判結果發現，易導致誤判之原因為特徵值落在臨界值邊緣之模糊地帶，基於此，本研究認為加入模糊理論的概念，應能降低誤判率，使整體的手勢辨識績效達到最佳。

A rule-based logical system was developed in the study for a recognition of 20 gestures of Chinese Sign Language. There are four tasks in the study: (1) motion tracking and 3D coordinates obtaining, (2) feature extraction, (3) rule-based recognition engine development, and (4) gesture recognition. Ten participants were recruited for the study. Data of the 5 participants were used for training and other 5 were used for gesture recognition. A total of 21 markers were attached onto the joints and fingertips of the hand of each participant. Vicon Motion Analysis Systems with 8 cameras was used to record the 20 static hand gestures and then the 3D coordinates of each marker were generated. Thirty-nine features of the hand (absolute distance of the markers and joint angles) were calculated. The features were then normalized according to the size of the hand of each individual. Rule-based recognition engine was developed using stepwise classification logic.
The results of study showed that merely seven eigenvalues were able to categorize the 20 gestures. They are: the absolute distance among middle finger, ring finger, and little finger(NADD3、NADD4、NADD5), the absolute distance between thumb and index finger, and thumb and middle finger(NRD1, NRD2), the curved angle of proximal interphalangeal joint between the finger tips of thumb and index finger(JAP1, JAP2). The best range of data for gestures recognition is to use the average feature value ± 2 time of standard deviation. Of the three engines, the model of ”p-relative location of thumb and index finger” out-performed “p-index finger stretch” and “p-little finger stretch, ” the best recognition rate of that model was 70%. Based on these experiences, a new rule-based model was developed and the recognition rate before any adjustment was 58%. It was recommended in the study that fuzzy logic should be applied to handle the harsh boundary values of each eigenvalues for improving the recognition rate.

1. Ali Erol, George Bebis, Mircea Nicolescu, Richard D. Boyle, (2005). Xander Twombly, A Review on Vision-Based Full DOF Hand Motion Estimation, Proceedings of the IEEE Workshop on Vision for Human-Computer Interaction (V4HCI), (in conjunction with the IEEE Conference on Computer Vision and Pattern Recognition), San Diego, California, June.
2. Azarbayejani, A., Wren, C., Pentland, A. (1996). Real-time 3-d tracking of the human body. Proceeding of IMAGE’COM 96, Bordeaux, France.
3. B. Stenger, A. Thayananthan, P. H. S. Torr, and R. Cipolla, (2003). “Filtering using a tree-based estimator,” Proc. IEEE Int. Conf. Computer Vision, pp. 1063-1070, Vol. 2.
4. B. Stenger, P. R. S. Mendonca, and R. Cipolla, (2001). “Model-based 3D tracking of an articulated hand,” Proc. IEEE Int. Conf. Computer Vision and Pattern Recognition, pp. 310-315, Vol. 2.
5. C. N¨olker and H. Ritter. (1999). Grefit: Visual recognition of hand postures. In A. Braffort, R. Gherbi, S. Gibet, J. Richardson, and D. Teil, editors, Gesture-Based Communication in Human-Computer Interaction: Proc. International Gesture Workshop, GW 99, France, pages 61–72. Springer Verlag, LNAI 1739.
6. C. S. Chua, H. Y. Guan, and Y. K. Ho. (2000). Model-based finger posture estimation. In ACCV2000.
7. C. Tomasi, S. Petrov, and A. Sastry, (2003). 3D tracking = classification + interpolation, Proc. IEEE Int. Conf. Computer Vision.
8. David J. Sturman, David Zeltzer, (1994). A Survey of Glove-based Input, IEEE Computer Graphics & Applications.
9. D.P. Huttenlocher, G.A.. Klandeman and W.J. Rucklidge (1993). Comparing images using the Hausdorff distance, IEEE Trans. On PAM I, vol.15, no. 9, pp. 850-863.
10. E. Holden. (1997). Visual Recognition of Hand Motion. PhD thesis, Department of Computer Science, University of Western Australia.
11. Eun-Jung Holden, Robyn Owens, Geoffrey G. Roy, (2000). An Adaptive Fuzzy Expert System for 3D Hand Motion Understanding, Computer Science and Software Engineering.
12. E. Oyama, A. Agah, Karl F. M., T. Maeda, S. Tachi, (2001). A modular neural network architecture for inverse kinematics model learning, Neurocomputing 40, 797-805.
13. Guangqi Ye, Jason J. Corso, Gregory D. Hager, (2004). Gesture Recognition Using 3D Appearance and Motion Features, CVPR Workshop on RTV4HCI.
14. Herv´e, J.-Y., (2000). Visual Hand Posture Tracking in a Gripper Guiding Application, International Conference on Robotics & Automation, 2, 1688-1694.
15. Iwai, Y., K. Watanabe, Y. Yagi and M. Yachoda, (1996). Gesture Recognition by Using Colored Gloves, IEEE International Conference on Systems, Man, and Cybernetics, 1:10, 76-81.
16. James Stewart, (1999). Calculus 4th,.
17. J. Davis and M. Shah, (1994). Recognizing hand gestures, Proc. ECCV’94, pp. 331-340.
18. Jintae Lee and T. L. Kunii, (1995). Model-based analysis of hand posture, IEEE Computer Graphics and Appl., vol.15, no.5, pp.77-86, Sep..
19. J. Lee and T. Kunii. (1993). Constraint-based hand animation. In Models and Techniques in Computer Animation, pages 110–127. Springer, Tokyo.
20. J. Triesch and C. von der Malsurg, (2001). A system for person-independent hand posture recognition against complex background, IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 23, No. 12, Dec.
21. J. Y. Lin, Y. Wu, and T. S. Huang, (2004). 3D model-based hand tracking using stochastic direct search method, Proc. IEEE Int. Conf. Automatic Face and Gesture Recognition, pp. 693-698.
22. K. Oka, Y. Sato, and H. Koike, (2002). Real-time fingertip tracking and gesture recognition, IEEE Computer Graphics & Applications, Vol. 22, No. 6, Nov..
23. L. Lee and T. L. Kunii, (1993). Constraint-based hand animation, Models and Techniques in Computer Animation, Tokyo: Springer-Verlag, pp.110-127.
24. Liu, F. Y. Zhuang, F. Wu, Y. P., (2003). 3D motion retrieval with motion index tree, Computer Vision and Understanding 92, 265-284.
25. Ma, J., Gao, W., Wu, J., and Wang, C. (2000). A continuous Chinese sign language recognition system. Proceedings of the Fourth, IEEE International Conference on Automatic Face and Gesture Recognition, pp.428-433.
26. M. Bray, E. Koller-Meier, and L. Van Gool, (2004). Smart particle filtering for 3D hand tracking, Proc. IEEE Int. Conf. Automatic Face and Gesture Recognition, pp. 675-680.
27. Mu-Chun Su, (2000). A Fuzzy Rule-Based Approach to Spatio-Temporal Hand Gesture Recognition, IEEE Transactions on Systems, Man, and Cybernetics—PartC: Applications and Reviews, vol. 30, No. 2, May.
28. Ng, C. and Ranganath, S., (2002). Real-time recognition system application, Image and Vision, 20, 993-1007.
29. N. Shimada and Y. Shirai, (1996). 3-D hand pose estimation and shape model refinement from a monocular image sequence, Proc. Of VSMM’96, pp. 423-428.
30. N. Shimada, K. Kimura, and Y. Shirai, (2001). Real-time 3-D hand posture estimation based on 2-D appearance retrieval using monocular camera, Proc. Int. Workshop RATFG-RTS, pp. 23-30.
31. O’Hagan, R.G., Zelinsky, A., Rougeaux, S., (2002). Visual gesture interfaces for virtual environment, Interacting with Computers, 14 , 231-250.
32. Park, S. W., Seo, Y., and Hong, K.S., (2000). Real-time camera calibration for virtual studio, Real-Time Imaging, 6, 433-448.
33. Rochelle O'Hagan, Alexander Zelinsky, (2002). Sebastien Rougeaux: Visual gesture interfaces for virtual environments. Interacting with Computers 14(3): 231-250.
34. Sato, Y., Kobayashi, Y., and Koike, H. (2000). Fast tracking of hands and fingertips in infrared images for augmented desk interface. IEEE Automatic Face and Gesture Recognition (FG2000), pp.462-467.
35. Segen, J., Kumar, S., (2000). Look ma, no mouse! , Communications of the ACM 43(7), 103-109.
36. V. Athitsos and S. Sclaroff, (2002). An appearance-based framework for 3D hand shape classification and camera viewpoint estimation, Proc. IEEE Int. Conf. Automatic Face and Gesture Recognition, pp. 45-50.
37. V. Athitsos and S. Sclaroff, (2003). Estimating 3D Hand Pose from a Cluttered Image, Proc. IEEE Int. Conf. Computer Vision and Pattern Recognition, pp. 432-439, Vol. 2.
38. V. Pavlovic, R. Sharma, and T. S. Huang, (1997). Visual Interpretation of Hand Gestures for Human-Computer Interaction: A Review, IEEE Trans. PAMI, vol. 19, No. 7, pp. 677-695, July.
39. Wang, X., (1999). A behavior-base inverse kinematics algorithm to predict arm prehension postures for computer-aided ergonomic evaluation, Journal of Biomechanics, 32, 453-460.
40. Y. Wu, J. Y. Lin, and T. S. Huang, (2001). Capturing natural hand articulation, Proc. IEEE Int. Conf. Computer Vision, pp. 426-432, Vol. 2.
41. Ying Wu and T. S. Huang, (2001). Hand modeling, analysis and recognition, IEEE Signal Processing, vol.18, no.3, pp.51-60, May.
42. Y. Yasumuro, Qian Chen, and K. Chihara, (1997). 3D modeling of human hand with motion constraints, Proc. of the Int. Conf. on 3-D Digital Imaging and Modeling, pp.275-282, May.
43. Yoon, H. S., Soh, J., Bae, Y. J., and Yang, H. S. (2001). Hand gesture recognition using combined features of location, angle and velocity. Pattern Recognition, 34(7), pp.1491-1501.
44. S.W. Lee, P.N. Braido, (2004). Towards an integrated high-fidelity linkage representation of the human skeletal system based on surface measurement, International Journal of Industrial Ergonomics: Special Issue on Anthropometrics and Disability, 33, pp. 215-227.
45. The 5DT Data Glove Driver Reference Manual, (2000).
46. 王天培 (1999) ：特徵手與類神經網路分類器於手勢辨識之研究，國立中山大學碩士論文。
47. 王國榮 (2001) 基於資料手套的智慧型手勢辨識之廣泛研究，國立臺灣科技大學碩士論文。
48. 許宏昌 (2003) 主成分分析法於指尖亮點手勢辨識之應用，國立中山大學海下技術研究所碩士論文。
49. 張良國 (2002) “基於Hausdorff 距離的手勢識別”, Journal of Image and Graphics. vol. 7, no. 11, pp. 1144-1150.
50. 陳治宇 (2003) 虛擬滑鼠：以視覺為基礎之手勢辨識，國立中山大學資訊工程學系碩士論文。
51. 陳偉政 (2003) 改良特徵抽取比對方法於手勢之識別，中華大學碩士論文。
52. 蘇木春, 張孝德, (2002) 機器學習：類神經網路、模糊系統以及基因演算法則, 全華科技圖書股份有限公司.
53. 財團法人資訊工業策進會資訊市場情報中心,(2003).新興科技介紹－人機介面技術, (2003). Advisory & Intelligence Service Program - IT Industry Overview, CDOC20030919003.