手部運動評估對於手部復健與治療的研究上具有重要意義，傳統上運動參數的取得是藉由量測黏貼於手指表面的標記點三維座標而得，可是在量測的過程中並未考慮標記點隨皮膚滑動的影響，而造成手指運動參數量測的失真，因此，皮膚與手指運動的關聯性是相當值得探討的主題，有鑒於此，我們藉由電腦視覺的技術追蹤及重建標記在手指上的網格，依據其結果並結合動態螢光攝影影像資訊驗證並量化皮膚滑動。
由於手部運動具有高自由度及限制性，因此系統結合多台相機降低遮蔽帶來的問題，本論文主要分成兩部分：預測並追蹤網格資訊，以及提出一套新的方法來量測皮膚滑動。在預測網格的部分我們採用了卡爾曼濾波器來預測網格的三維資訊，並且將此投影到影像平面上，提供追蹤時的初始位置，追蹤的部分則是結合Mean-shift演算法與Dual template matching並結合結構性資訊來定位出每個時間點的二維網格，選擇可靠點重構出三維網格；為了驗證皮膚滑動的存在並量化，量測滑動程度的部分我們藉由手動對位求出骨頭座標系的轉換關係，即可求出在不同時間點的未滑動點座標，並與重構網格投影在動態螢光攝影平面的滑動點比較，藉此定義出位移向量。
根據效能實驗顯示出系統在二維追蹤上有著良好的追蹤效果，我們也可根據重構出的網格量測到手指關節運動時，手部表面面積的變化，並結合動態螢光攝影獲得皮膚滑動情形。藉此分析病患手部運動時皮膚的形變情況是否有所異常，進而建立更完整的病歷資訊。

In hand rehabilitation and treatment, kinematic assessment of hand plays an important role. Traditionally, the kinematic parameters were measured by the 3D coordinates of markers attached to the skin surface, but the process of measurement doesn’t include the effect of skin movement with respect to the bone. Therefore, the relationship between skin and finger motion is worthwhile to research. In this thesis, we have developed an analysis system to measure the extent of skin movement by utilizing the computer vision techniques to reconstruct the grid attached to the finger. According to the result, the validation and quantization of skin movement can be done by utilizing both the video and fluoroscopy to capture the image information of muscular skeletal motion.
Since the finger motion is with high degrees of freedom and also highly constrained. Therefore, our system combines multiple cameras to deteriorate the effect of occlusion, and increase the accuracy of reconstruction simultaneously. The thesis consists of two parts: predicting and tracking the grids and providing a new method to measure the skin movement. In the grid prediction, we apply the Kalman filter to predict 3D grid, and project it back to the image plane. The projected grid provides the initial position of tracking process. At the tracking stage, we utilize the mean-shift algorithm and dual template matching to locate the 2D grid in motion. In measuring the skin movement, we use the manual registration to calculate the transformation of BCS(Bone coordinate system). Then, we can get the coordinate of markers without skin movement by the transformation at different frame. The offset vector of skin movement is defined by the points with and without skin movement, and the point with skin movement is calculated by the projection of reconstructed grid on the projection plane of fluoroscopy.
According to the performance experiment, the system has good tracking result. We can measure the deformation extent of finger surface from the reconstructed 3D grid, and find the skin movement by comparing with markers in fluoroscopy images. The system can analyze whether the skin deformation is normal or not when the patient moves finger motion. The more detailed case report on finger motion can be created based on these measurements.

[1]Li-Chieh Kuo, William P. Cooney III , Mineo Oyama, Kenton R. Kaufman,Fong-Chin Su, Kai-Nan An, “Feasibility of using surface markers for assessing motion of the thumb trapeziometacarpal joint,” Clinical Biomechanics,18:558-563,2003
[2] Wen-Chieh Lin, Yanxi Liu, “A Lattice-Based MRF Model for Dynamic Near-Regular Texture Tracking,” IEEE Trans. PAMI29(5)(2007) 777 – 792
[3] Arulampalam, M.S. Maskell, S. Gordon, N. Clapp, T. ,“A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking,” Signal Processing, IEEE Transactions , vol. 50, iss.2, pp. 174-188, Feb 2002.
[4] Baker,S., and Matthews,I.,“Lucas-Kanade 20 Years On: A Unifying Framework,” International Journal of Computer Vision, 53(3), pp221–255, 2004.
[5] B Liu, R Manner, “An Efficient and Accurate Method for 3D-Point Reconstruction from Multiple Views,” International Journal of Computer Vision 65(3), 175–188, 2005.
[6]Dorin Comaniciu, and Peter Meer, “Mean Shift: A Robust Approach Toward Feature Space Analysis,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 24, no. 5, pp. 603-619, May 2002.
[7] D. Comaniciu, V. Ramesh, and P. Meer, “Kernel-Based Object Tracking,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 25, no. 5, pp. 564-575, May 2003.
[8] B Liu, R Manner, “An Efficient and Accurate Method for 3D-Point Reconstruction from Multiple Views, “International Journal of Computer Vision 65(3), 175–188, 2005.
[9] M. Zaki, M. Y. El Nahas and M. Youssef, “EMBOT: An enhanced motion-based
object tracker,” The Journal of Systems and Software, vol. 69, no. 1-2, pp.
149-158, Jan. 2004
[10] G. Rigoll, H. Breit and F. Wallhoff, “Robust tracking of persons in real-world
scenarios using a statistical computer vision approach,” Image and Vision
Computing, vol. 22, no. 7, pp. 571-582. July 2004.
[11] G. Welch and G. Bishop, “An introduction to the kalman filter,” Technical
Report TR 95-041, University of North Carolina, Department of Computer
Science, 1995.
[12] Cappozzo, A., Cappello, A., Croce, U.D., Pensalfini, F., 1997. “Surface-marker cluster design criteria for 3-D bone movement reconstruction,” IEEE Transaction on Biomedical Engineering 44(12), 1165–1174.
[13] Lucchetti, L., Cappozzo, A., Cappello, A., Croce, U.D., 1998. “Skin movement artifact assessment and compensation in the estimation of knee joint kinematics,” Journal of Biomechanics 31, 984–997.
[14] Alexander, E.J., Andriacchi, T.P., 2001. “Correcting for deformation in skin-based marker systems,”Journal of Biomechanics 34 (3),355–361.
[15]Jae Hun Ryu, Natsuki Miyata, Makiko Kouchi, Masaaki Mochimaru, Kwan H. Lee, “Analysis of skin movement with respect to flexional bone motion using MR images of a hand” Journal of Biomechanics 39,(2006) 844 - 852.
[16] E.H. Garling, B.L. Kaptein, B. Mertens, W. Barendregt, H.E.J. Veeger,
R.G.H.H. Nelissen, E.R. Valstar “Soft-tissue artefact assessment during step-up using fluoroscopy and skin-mounted markers,” Journal of Biomechanics 40,(2007) S18 - S24
[17] 陳彥廷、孫永年，“以模型為基礎之電腦手指連續運動影像分析系統”，國立成功大學資訊工程研究所碩士論文，2006。
[18]. 何聖斌、孫永年，“使用多台相機的三維手指影像運動分析系統”，國立成功大學資訊工程研究所碩士論文，2007。
[19]http://en.wikipedia.org/wiki/Image:Scheme_human_hand_bones-en.svg
[20]http://en.wikipedia.org/wiki/Image:Gray337.png
[21]http://en.wikipedia.org/wiki/Hinge_joint
Wen-Chieh Lin, Yanxi Liu, “A Lattice-Based MRF Model for Dynamic Near-Regular Texture Tracking,” IEEE Trans. PAMI29(5)(2007) 777 – 792
[22]Igor Guskov, Sergey Klibanov, Benjamin Bryant, “Trackable surfaces,” SIGGRAPH(2003)
[23]Ryan White, Keenan Crane, D.A. Forsyth, “Capturing and Animating Occluded Cloth,” SIGGRAPH(2007)