臺灣博碩士論文加值系統

摘要:
良好的數位視訊索引建立技術(video indexing)以完成之視訊資料庫系統對方便視訊之瀏覽(Browsing)及擷取(retrieval)是相當必要而且具關鍵性因素的，在建立視訊索引之系統中在視訊中偵測出不同的視訊鏡頭(shot)與場景(scene)來作分析，以萃取出低階特徵來建立視訊索引之系統是現今所知具相當有價值也是具挑戰性的研究；而在視訊鏡頭偵測對於監控系統或建立視訊資料庫等應用是十分重要之前置處理，在本論文最關鍵的部分是我們將三階層式的視訊鏡頭偵測系統做為核心，並使用區域輪廓套索特徵值擷取演算法來針對剛體之形變的漸溶、漸變(Gradual, Dissolve)的場景做有效的偵測。
本論文所提之視訊偵測系統，我們第一階段提出一個鏡頭(shot)及場景(scene)之三層偵測系統作視訊特性特徵偵測，其第一層試圖以整張影像當視訊的物件，本論文初步已定義出每張影像邊緣均衡分割中心，經模擬已可以大致快速追蹤到整張影像變化趨勢，偵測系統之第二層，透過邊緣均衡分割中心將影像再分四小區，做顏色分佈之變化趨勢追蹤，偵測系統之第三層利用輪廓套索特徵值擷取影像分割追蹤之演算法，由主要變化之兩張影像內容分出主物件，並作相似度比對經由曲率值的擷取來做相似度分析以便可以針對漸進式形變做偵測，以上系統階段就是完成鏡頭類型、重要性和內容之偵測及特徵擷取；就整體架構來說上層的部分可以初步偵測出具體明顯的形變而到了下層我們可以更進一步偵測特殊的形變，根據各層處理過程及特徵本身特性來針對不同的視訊鏡頭變化類型做實際的分類偵測；在將來我們可以將此架構運用在監測系統上或物件的運動追蹤等[16]。

Abstract:
It is an effective technique for the video browsing and retrieving to construct the video indexing of digital video database. In the video indexing system, detecting be different the shots and scene for analysis from video sequence, the efficient extraction of low-level features from the shots in the video string is very valuable and the research of challenge; it is important the preprocesses for surveillance or video indexing system at video shots detection system. In this thesis, the three-layer shot detection system will be kernel of system. At third-layer, the region contour feature be used to detect shots of gradual or dissolve.
This thesis proposal attempts to develop a three-layer shot detection system as the first stage of video indexing system, which contains three layers. a frame is regarded as a object member of a video in the first detection layer, we have already found a balance splitting center for each frame so far and can exploit it to roughly tracing the shot changes. The second layer uses four specified regions, which are segmented by the balance splitting center to yield more precise shot changes via the histogram-based colorific analysis and comparison. Then the third layer extracts the major objects from the obviously changed frames based on the region contour feature algorithm. By the similarity comparison of these curvature values of the objects, the detailed detection and the type decision of shots can be achieved. As for proposed architecture could detect obviously deformation of object on top-layer, then we can further detecting deformation of gradual on bottom-layer, According to our results of layer processed and feature of video sequence for classification, in the feature, this architecture will be applied to video surveillance or object tracking.

摘要: I
ABSTRACT: III
致 謝 V
目 錄 VI
圖 表 目 錄 IX
第1章 1
緒 論 1
1.1 問題描述: 1
1.2 研究背景: 2
1.3 研究動機與目的: 3
第2章 4
相關文獻研究回顧 4
2.1 VIDEO SHOTS的分類探討: 4
2.2傳統偵測場景變化(SCENE CHANGE)的方法: 7
2.3直方圖偵測準則 (HISTOGRAM CRITICAL): 8
第3章 9
三層式架構視訊偵測相關研究與分析 9
3.1 三階層式視訊鏡頭偵測研究方法與進行步驟: 9
3.2 三階層式視訊鏡頭偵測系統基本示意圖: 10
3.3 三階層式視訊鏡頭偵測演算法： 10
3.3-1.第一層次偵測： 11
3.3-1-1視訊鏡頭偵測第一層運作步驟: 12
3.3-1-2 影像序列進入第二層判斷的條件為: 15
3.3-2.第二層次偵測： 15
3.3-2-1視訊鏡頭偵測第二層運作步驟: 16
3.3-2-2影像序列進入第三層判斷的條件為: 19
3.3-3.第三層次偵測： 19
3.3-4. 三層式視訊鏡頭偵測系統流程圖: 19
第4章 21
以局部特徵區域演算法針對漸進式視訊形變偵測之分析 21
4.1 偵測漸變及溶解運鏡變化演算法乃用於決定層，分四大過程(子演算法)： 21
4.1-1子演算法第一步驟： 21
4.1-2子演算法第二步驟： 21
4.1-3子演算法第三步驟： 22
4.1-4子演算法第四步驟： 22
4.2 用於三階層式中末層-決定層-偵測漸變及溶解運鏡變化演算法： 23
4.2-1 特徵值取點演算法: 23
4.2-2 初步判定序列影像變化: 23
4.2-3 三階層式視訊運鏡偵測漸融場景之步驟: 24
4.2-4 三階層式視訊運鏡偵測漸融場景之演算法則: 26
4.3 輪廓套索特徵值擷取演算法介紹: 29
第5章 33
實驗結果與討論 33
5.1 三階層式視訊運鏡偵測系統之結果: 33
5.2三階層式視訊運鏡偵測漸融場景之實驗結果: 37
5.2-1物件位移式的序列影像模擬步驟: 37
5.2-2漸變式的序列影像模擬步驟: 41
5.3偵測效能評估: 43
第6章 49
結論與未來方向 49
結論未來展望論述: 49
參 考 文 獻 50

[1] C. O''Toole, A. Smeaton, N. Murphy and S. Marlow, “Evaluation of Automatic Shot Boundary Detection on a Large Video Test Suite”, The Challenge of Image Retrieval (CIR 99) - 2nd UK Conference on Image Retrieval. Newcastle, UK, 25-26, February, 1999.
[2] J. H. Oh, K. A. Hua, N. Liang, “A content-based scene change Detection and classification technique using background tracking”, Proc. of IS&T/SPIE conference on Multimedia Computing and Networking 2000. San Jose CA. pp. 254-265 Jan. 24 - 28, 2000.
[3] K. A. Hua, J. H. Oh and K. Vu,” Non-Linear Approach to Shot Boundary Detection”, IS&T/SPIE Conference on ultimedia Computing and Networking 2001. San Jose CA. pp. 1-12. Jan. 22-25, 2001.
[4] P. Browne, A. Smeaton, N Murphy, N. O''Connor, S. Marlow and C. Berrut “Evaluating and Combining Digital Video Shot Boundary Detection Algorithms”, Irish Machine Vision and ImageProcessing Conference (IMVIP 2000), Belfast, Northern Ireland, 31 August - 2 September 2000.
[5] R. Zabih, J. Miller, K. Mai, “A feature-based algorithm for detecting and classifying production effects”, Multimedia Systems Computer science dept. Cornell Univ, Ithaca, NY, 119-128, 7 (1999).
[6] X. Liu and T. Chen, “Shot Boundary Detection Using Temporal Statistics Modeling”, IEEE International Conference on Acoustics, Speech, and Signal Processing, 2002. Proceedings. (ICASSP ''02). Vol.4 , 13-17 pp. IV-3389 -IV-3392 May 2002
[7]. X. Cao; Suganthan, P.N.;Neural Networks, “Hierarchical overlapped growing neural gas networks with applications to video shot detection and motion characterization”, 2002. IJCNN ''02. Proceedings of the 2002 International Joint Conference on , Vol. 2, 12-17, pp.1069 —1074, May 2002.
[8] X. Gao and X. Tang, “Unsupervised video shot segmentation and model-free anchorperson detection for news video story parsing,” IEEE Transaction on Circuits, Systems and Video Technology, Vol. 12, no. 9, pp. 765 —776, Sept., 2002.
[9] S. Satoh, “News video analysis based on identical shot detection”,
In Proc. 2002 IEEE Intl. Conf. On Multimedia and Expo, Vol. 1, pp. 692, Aug.2002.
[10] Bruno, E.; Pellerin, D, “ Video shot detection based on linear prediction of motion”, IEEE International Conference on Multimedia and Expo, 2002. ICME ''02, Proceedings. 2002 , Vol. 1 , 26-29, pp. 289 -292 Vol.1, Aug. 2002.
[11] Z. Cernekova, C. Nikou, I. Pitas, “Shot detection in video sequence using entropy-based metrics,” 2002 International Conference on Image Processing., Vol. 3, pp. 421 -424, 2002.
[12] N. V. Patel and I. K. Sethi, “Compressed video processing for cut detection”, Vision IEE Proceedings- Image and Signal Processing, Vol. 143 Issue: 5, pp. 315 -323, Oct 1996.
[13]. K. Sethi and N. V. Patel, “A statistical approach to scene change detection”, in IS&T SPIE Proceedings: Storage and Retrieval for Image and Video Databases IIII, Vol. 2420, (San Jose), pp. 329-- 339, February 1995.
[14] Y. Rui, T. S. Huang and S. Mehrotra, “Exploring video structure beyond the shots”, IEEE International Conference on Multimedia Computing and Systems, pp. 237 -240, 28 Jun-1 Jul 1998.
[15] W. Zhao, J. Wang, D. Bhat, K. Sakiewicz, N.Nandhakumar, “Improving color based video shot detection”, IEEE International Conference on Multimedia Computing and Systems, Vol. 2, pp. 752 -756, Jul. 1999 .
[16] Stringa, E.; Regazzoni, C.S., “Real-time video-shot detection for scene surveillance applications”, IEEE Trans. on Image Processing , Vol. 9 Issue: 1 , pp. 69 -79 , Jan. 2000.
[17]. M. K. Andrew Witkin and D. Terzopoulos, “Snakes: Active C Contour Models”, International Journal of computer vision, Kiuwer academic publishers Boston manufactured in the Netherland, 1987.
[18]. B. Bascle and R. Deriche, “Features Extraction Using Prametric Snakes”,Pattern Recognition, 1992 Conference C: Image, Speech and Signal Analysis, Proceedings, 11th IAPR International Conference, Vol.3, Sept. 1992.
[19]. L. Ji, and H. Yan,”Loop-free Sankes for Image Segmentation”,
1999 Image Processing, , ICIP99 Proceedings, Vol. 3, pp.24-28 Oct. 1999.
[20]. L. Ji, and H. Yan, “Robust Topology-Adaptive Snakes for Image Segmentation”,
2001 Image Processing, ICIP2001 Proceedings, Vol. 2, pp.797 —800, Oct. 2001.
[21]. S. W. Jang; El-Kwae, E.A.; Hyung-Il Choi, “Shaking Snakes Using Color Edges for Contour Extraction”, 2002 Image processing,ICIP2002Proceedings, Vol.2 , pp.817-820, Sept. 2002.
[22] Student: W. L. Wu, Advisor: Dr. R. Feng, “Segmentation of Breast Tumor in 3D Ultrasound image using discrete Active Contour Model”, Chang, Institute of Computer Science and Information Engineering, Nation Chung-Chang University, Taiwan. 2000.
[23] Erkmen, I.; Erkmen, A.M.; Matsuno, F.; Chatterjee, R.; Kamegawa, T., “Snake robots to the rescue”, IEEE , Robotics & Automation Magazine, Vol. 9 Issue: 3 , pp. 17 —25, Sept. 2002.
[24] S. Sun; Haynor, D.R.; Y. Kim, “Semiautomatic video object segmentation using Vsnakes”, IEEE Transactions on Circuits and Systems for Video Technology , Vol. 13 Issue: 1 , pp. 75 —82, Jan. 2003.
[25] Davatzikos, C.A.; Prince, J.L., “An active contour model for mapping the cortex”, IEEE Transactions on Medical Imaging, Vol. 14 Issue: 1 , pp. 65 —80, March 1995.
[26] In the Compressed Domain:
U. Gargi, R. Kasturi, Susan H. Strayer. “Performance Characterization of Video-Shot-Change Detection Methods.” IEEE Transaction on Circuits and Systems for Video Technology, Vol. 10, No. 1, February 2000.
[27] In the Uncompressed Domain:
Z. Cernekova, C. Nikou, I. Pitas. “Shot Detection in Video Sequences Using Entropy-Based Metrics.” International Conference on Image Processing 2002 (ICIP2002), Vol. 3, pp. 421-424, 2002.
[28] Dissolve Detection
J. Nam and A. H. Tewfik. “Dissolve Transition Detection Using B-Splines Interpolation.” IEEE International Conference on Multimedia and Expo (ICME), July 2000.
[29] Wipe Detection
J. Nam and A. H. Tewfik.” Wipe Transition Detection Using Polynomial Interpolation.” In Storage and Retrieval for Media Databases 2001, Proc. SPIE 4315, pp. 231-241, Jan. 2001.
[30] Comparative Surveys:
Y. Gong, X. Liu “Video shot segmentation and classification”, C&C Research LAB. NEC USA, INC
[31] Comparative Surveys:
U. Gargi, R. Kasturi, and S. Antani. “Performance Characterization and Comparison of Video Indexing Algorithms”. Proc. IEEE Conference on Computer Vision and Pattern Recognition, Santa Barbara, CA, pp. 559-565, June 1998.
[32] B. L. Yeo; B. Liu. “Rapid scene analysis on compressed video” Circuits and Systems for Video Technology, IEEE Transactions on , Vol. 5 , Issue: 6 , pp. 533 — 544 Dec. 1995.