運動軌跡（motion track）是視訊資料中一個重要的特徵值，它可以表現運動物體在時間與空間上的相對關係，也是內涵式影片查詢中一項重要的索引。擷取運動軌跡的方法，就是將物體在每個畫面的位置連接起來，在壓縮視訊資料內可利用運動向量（motion vector）可推導出來。為了能更充分利用及管理運動軌跡的資訊，在這篇論文中，我們在MPEG-7的運動描述子（motion descriptor）之上，提出了一個新的運動軌跡表示方法，加入了在X-Y平面上表示軌跡的描述，以及在每個軸上對速度趨勢的描述。我們使用一次、二次或三次的多項式來作運動軌跡的表示方法，並使用多項式的係數作為特徵值。在這樣的表示方法之上，我們會定義新的相似度測量方法，分別是在軌跡方面，根據不同次方的曲線的不同特性，例如局部最大/最小值、平滑及延展程度等等，訂定其相似特徵值，以及在速度方面根據不同的速度趨勢變化來訂定相似的程度。我們更進一步提出分段的方法，以方便在這樣的表示法上處理有複雜行為的運動軌跡，並提出索引結構及比對的方法以進行有效率的進行查詢處理。在查詢處理之後，使用相關回饋（relevance feedback）的技巧來調整相似度各個特徵值距離的比重，使用者可以藉由回饋的機制，得到他所想要的查詢結果。實驗結果顯示我們的方法可以達到約85%的精準度，同時在查詢處理上是有效率的，並可以作多段運動軌跡的查詢。

The motion track is an important feature to show the spatio-temporal relationship of a video object. In this thesis, we propose a novel motion track representation based on MPEG-7 motion descriptor. A new descriptor is proposed to represent the motion track in the X-Y plane and the trend of velocity changes. We use the 1st/ 2nd/ 3rd order polynomials to model the trajectory of moving objects and use the associated coefficients as their features. Moreover, a new similarity measure for comparing two motion tracks based on the motion trajectory and velocity differences is proposed. The trajectory is compared by the properties of the polynomials such as the peaks and tangent lines, and the velocity is compared by the different trends such as speeding up and speeding down. Furthermore, the motion track segmentation method is proposed to handle a complicated motion behavior. The indexing and matching algorithm are also considered and the relevance feedback is used to improve the query results. The experiments show the query processing is efficient and the precision is near 85%, which is higher than other approaches.

[1] “Overview of the MPEG-7 Standard,” Dec. 2001, http://mpeg.cselt.it/.
[2] S.F. Chang, W. Chen, H. Meng, H. Sundaram, D. Zhong, “A fully automated content-based video search engine supporting spatiotemporal queries,” IEEE Transaction on Circuit Systems Video Technology, Volume:8, Issue:5, p 602 -615 , 1998.
[3] S.F. Chang, W. Chen, H. Meng, H. Sundaram, and D. Zhong, “VideoQ: an automated content based video search system using visual cues” Proceedings of ACM Multimedia ‘97, Nov. 1997.
[4] William Chen and Shih-Fu Chang, “Motion Trajectory Matching of Video Objects,” IS&T/SPIE's 13th Annual International Symposium (SPIE), 2000.
[5] Serhan Dagtas, Wasfi Al-Khatib, Arif Ghafoor, Ashfaq Khokhar, “Trail-Based Approach for Video Data Indexing and Retrieval,” International Conference on Multimedia Computing and Systems (ICMCS), 1999.
[6] Serhan Dagtas, Wasfi Al-Khatib, Arif Ghafoor, and Rangasami L. Kashyap, “Models for Motion-Based Video Indexing and Retrieval,” IEEE Transaction on Image Processing, Volume:9, Issue:1, p 88 -101 Jan.2000.
[7] Yoshiharu Ishikawa, Ravishankar Subramanya, Christos Faloutsos, “MindReader: Querying databases through multiple examples,” Proc. 24th Int. Conf. Very Large Data Bases (VLDB), 1998.
[8] Sylvie Jeannin and Ajay Divakaran, “MPEG-7 visual motion descriptors,” IEEE Transactions on Circuits and Systems for Video Technology, Volume: 11, Issue: 6 , June 2001.
[9] Sylvie Jeannin, Radu Jasinschi, Alfred She, Thumpudi Naveen, Benoit Mory, Ali Tabatabai, “Motion descriptors for content-based video representation,” Signal Processing: Image Communication 16,p 59-85, 2000.
[10] T. Kahveci and A. Singh, “Variable Length Queries for Time Series Data,” International Conference on Data Engineering (ICDE), 2001.
[11] Eamonn J. Keogh, Michael J. Pazzani, “Relevance Feedback Retrieval of Time Series Data,” SIGIR 1999.
[12] Eamonn Keogh, Kaushik Chakrabarti, Sharad Mehrotra, Michael Pazzani, “Locally Adaptive Dimensionality Reduction for Indexing Large Time Series Databases,” ACM International Conference on Management of Data (SIGMOD), 2001
[13] Tony C. T. Kuo and Arbee L. P. Chen, “Content-Based Processing for Video Databases,” IEEE Transaction on Multimedia, Vol. 2, No. 1, pp. 1-13, 2000.
[14] Tony C. T. Kuo, Y. B. Lin and Arbee L. P. Chen et al, “Efficient Shot Change Detection on Compressed Video Data,” Proc. of IEEE Workshop on Multimedia Database Management systems, August 1996, pp. 101-108.
[15] KyuWon Lee, Woong-Shik You and Jinwoong Kim, “Video Retrieval based on the Object’s Motion Trajectory,” Visual Communications and Image Processing (VCIP), 2000.
[16] John Z. Li, Tamer Ozsu, Duane Szafron, “Modeling of Moving Objects in a Video Database,” International Conference on Multimedia Computing and Systems (ICMCS), 1997.
[17] C. H. Lin and Arbee L. P. Chen, “Motion Event Derivation and Query Language for Video Databases,” in Storage and Retrieval for Media Databases 2001, Proceedings of SPIE Vol. 4315, pp.208-218.
[18] C. C. Liu, Jia-Lien Hsu, Arbee L. P. Chen, “An Approximate String Matching Algorithm for Content-Based Music Data Retrieval,” International Conference on Multimedia Computing and Systems (ICMCS), 1999.
[19] C. C. Liu and Arbee L. P. Chen, “Vega: A Multimedia Database System Supporting Content-Based Retrieval,” Journal of Information Science and Engineering, Vol. 13, No. 3, Sep. 1997, pp. 369-398.
[20] C. C. Liu and Arbee L. P. Chen, “3D-List: A Data Structure for Efficient Video Query Processing,” IEEE Trans. on Knowledge and Data Engineering, Volume: 14, Issue: 1, pp. 106-122 , 2002.
[21] Ye Lu, Chunhui Hu, Xingquan Zhu, HongJiang Zhang, Qiang Yang, “A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems,” ACM Multimedia 2000.
[22] Polly Wan Po Man, Man Hon Wong, “Efficient and Robust Feature Extraction and Pattern Matching of Time Series by Lattice Structure,” The Chinese University of Hong Kong, ACM International Conference on Information and Knowledge Management (CIKM), 2001
[23] Yang-Sae Moon, Kyu-Young Whang, and Woong-Kee Loh, “Duality-Based Subsequence Matching in Time-Series Databases,” International Conference on Data Engineering (ICDE), 2001.
[24] Jianhao Meng, Yujen Juan, and Shih-Fu Chang, “Scene change detection in a mpeg compressed video sequence,” SPIE Electronic Imaging: Science & Technology- Digital Video Compression: Algorithms and Technologies, 1995.
[25] Eitetsu Oomoto and Katsumi Tanaka, “OVID: Design and Implementation of a Video-Object Database System,” IEEE Transactions on Knowledge and Data Engineering, August 1993, pp. 629-643.
[26] Chang-Shing Perng, Haixun Wang, Sylvia R. Zhang, D. Stott Parker, “Landmarks: A New Model for Similarity-Based Pattern Querying in Time Series Databases,” International Conference on Data Engineering (ICDE) 2000
[27] Joseph John Rocchio, “Relevance feedback in information retrieval,” The SMART Retrieval System: Experiments in Automatic Document Processing, pp. 313-323. Prentice Hall, 1971.
[28] Yong Rui, Thomas Huang, “Optimizing Learning in Image Retrieval,” IEEE conference on Computer Vision and Pattern Recognition 2000.
[29] Yong Rui, Thomas S. Huang, and Sharad Mehrotra. “Human perception subjectivity and relevance feedback in multimedia information retrieval.” In Proceedings of IS&T and SPIE Storage and Retrieval of Image and Video Databases VI, 1998.
[30] Choon-Bo Shim and Jae-Woo Chang, “Spatio-temporal Representation and Retrieval Using Moving Object's Trajectories,” ACM Multimedia Workshop, 2001.
[31] Ting-Yu Wai, Arbee L. P. Chen, “Retrieving Video Data via Motion Tracks of Content Symbols,” ACM International Conference on Information and Knowledge Management (CIKM), 1997.
[32] Atsuo Yoshitaka, Masanori Yoshimitsu, Masahito Hirakawa, Tadao Ichikawa, “V-QBE: Video Database Retrieval by Means of Example Motion of Objects,” International Conference on Multimedia Computing and Systems (ICMCS), 1996.
[33] Woongshik You, Kyu Won Lee, Jae-Gon Kim, Jimwoong Kim and Oh-Seok Kwon, “Content-Based Video Retrieval by Indexing Objects Motion Trajectory,” International Conference on Computers in Education (ICCE), 2001.