跳到主要內容

臺灣博碩士論文加值系統

(54.224.117.125) 您好!臺灣時間:2022/01/23 20:23
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:王璟勳
研究生(外文):Ching-Hsung Wang
論文名稱:內容導向MPEG-4網格物件之建立
論文名稱(外文):The Generation of Content-basedMesh Object in MPEG-4
指導教授:陳進興陳進興引用關係
指導教授(外文):Jin-Xing Chen
學位類別:碩士
校院名稱:國立成功大學
系所名稱:電機工程學系碩博士班
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2002
畢業學年度:90
語文別:英文
論文頁數:54
中文關鍵詞:視訊壓縮Mesh物件
外文關鍵詞:mesh objectMPEG-4
相關次數:
  • 被引用被引用:1
  • 點閱點閱:117
  • 評分評分:
  • 下載下載:13
  • 收藏至我的研究室書目清單書目收藏:0
視訊壓縮標準MEPG-1, MPEG-2, H.261, 及 H.263,均以方塊比對作運動預測及補償來去除時間上的資訊重複,但在低位元率的情況下,方塊比對的方法容易造成所謂的方格效應。為了解決這個問題,MPEG-4在視訊物件中,制定了網格物件,網格物件編碼以更廣義的空間轉換warping來對物件中的每個patch作運動補償。
本論文提出一網格產生的方法,網格包含了兩種,一是intra網格、另一是inter網格。intra網格由節點及節點所形成的patches所構成。Inter網格則由每一節點的運度向量所決定。Intra網格的節點包括兩種,一在物件邊界上,一物件內部裡,本論文方法以高曲度變化及高灰度變化分別作為選取前者及後者的依據。為了避免網格交疊,本論文移除一些非必要的特徵點,然後根據這些特徵點,利用Delaunay三角化法求出網格上的所有三角形patches。對於inter網格本論文提出一兩層式(粗糙和精細)搜尋方法來預測節點運動向量。在粗糙這一層中,本論文以方塊比對求出最佳運動向量,在精細這一層,本論文以affine transformation來決定最佳的運動向量。
本論文的實驗包括兩組不同設定值,一組節點數較少、另一組節點數較多,實驗結果顯示前者的PSNR平均值為17.4,後者的PSNR平均值為21.3。與其他的編碼方法比較,本論文方法的計算較簡單,而影像品質則相近。
The block matching method used by MPEG-1, -2/H.261, H.263 for motion prediction/compensation suffers from blocking artifacts at low bit rates. To avoid such a problem, a new coding method in which a mesh object is encoded was developed in MPEG-4. In the mesh-based coding method, motion compensation is accomplished by a more general spatial transformation (warping) related to the geometry of the mesh element.
This thesis proposed a method for mesh generation. There are two types of mesh to be considered- one is the intra mesh and the other is the inter mesh. The intra mesh is represented as a collection of patches formed by the node points. The inter mesh pertains the motion vector of each node point. In the intra mesh, two kinds of nodes are considered: one is on the boundary, and the other is inside the object. Our proposed method extracts feature points of high curvature magnitude and high edge magnitude for the former and the latter repectively. To avoid crossover among the mesh patches, some undesired feature points are removed. The feature points extracted are then triangulated by the Delaunay triangulation method to get the mesh. In the inter mesh, a two-level (coarse and refinement) scheme is proposed to predict the motion vectors of the node points. In the coarse level, the block matching algorithm (BMA) is used to provide an initial estimate of the motion vector. In the refinement level, the affine mapping is used to obtain the refinement of the motion vector.
Two cases of different setting conditions were tested in the experiment. One has fewer node points than the other. The average PSNR of the former is 17.4 and that of the latter is 21.3. Compared with other methods, the proposed method is faster and the image quality is about the same.
中文摘要 i
英文摘要 ii
CONTENTS vi
List of Tables xi
List of Figures xii
Chapter 1 Introduction 1
1.1 Motivation 1
1.2 Recent Works 2
1.3 Thesis Organization 3
Chapter 2 Mesh Representation of a Video Object 4
2.1 Definition of Mesh 4
2.2 Mesh Generation 5
2.2.1 Image Preprocessing 6
2.2.2 Spatial Transformation for Image Warping 7
2.2.3 Computational Geometry 9
2.3 Applications for Mesh 10
Chapter 3 Mesh Object of MPEG-4 12
3.1 Visual Objects 12
3.2 Mesh Object 13
3.3 Mesh Object Planes 15
3.4 The Syntax and Semantics of the Mesh Object 16
3.5 Discussions 27
Chapter 4 A Content-Based Method for Mesh Generation 28
4.1 Flow of Mesh Generation 28
4.2 Intra Mesh Generation 30
4.2.1 Generation of Boundary Node Points 31
4.2.1.1 Object Shape Representation Using Chain Codes 31
4.2.1.2 Feature Selection Based on Polygonal Approximation 31
4.2.2 Generation of Interior Node Points 33
4.2.2.1 Edge Detection 33
4.2.2.2 Node Points Generation by Quadtree Splitting 35
4.2.2.3 Removal of Interior Node Points Outside the Approximation Polygon 36
4.2.3 Undesired Feature Points Removal 37
4.2.4 Delaunay Triangulation 38
4.3 Inter Mesh Generation 38
4.3.1 Coarse Motion Estimation Using Block Matching 39
4.3.1.1 Block Matching Using Logarithmic Search 40
4.3.2 Refinement of Motion Estimation Based on Affine
Transformation 41
4.3.2.1 Generation of Bounding Polygons 41
4.3.2.2 Affine Transformation 42
4.3.2.3 Generation of Candidate Points 43
Chapter 5 Experiment Results and Conclusion 46
5.1 Simulation Results 46
5.2 Discussion 49
5.3 Conclusions and Future Work 54
[1]ISO/IEC JTC1/SC29/WG N2502a, ISO/IEC 14496-2 Visual, Atlantic City, October 1998.
[2]Y. Nakaya and H. Harashima, “Motion Compensation Based on Spatial Transformations,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 4, No. 3, pp. 339-367, June 1994.
[3]A. M. Tekalp, P. V. Beek, N. Zhuang, I. Celasun and M. Xia, “Hierarchical 2-D Mesh Representation, Tracking and Compression for Object-Based Video,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 9, No. 2, pp. 353-369, March 1999.
[4]A. M. Tekalp and I. Celasun, “Optimal 2-D Hierarchical Content-Based Mesh Design and Update for Object-Based Video,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 10, No. 7, pp. 1135-1153, October 2000.
[5]A. M. Tekalp, I. Celasun, E. Ilgaz, P. V. Beek and N. Zhuang, “Optimal Hierarchical Design of 2D Dynamic Meshes for Videos,” Image Processing, ICIP 98. Proceedings, Vol. 8, pp. 899-903, 1998
[6]A. M. Tekalp, M. I. Sezan and C. Toklu, “Tracking Motion and Intensity Variations Using Hierarchical 2-D Mesh Modeling for Synthetic Object Transfiguration,” Graphical Models and Image Processing, Vol. 58, No. 6, pp. 553-573, November 1996.
[7]A. M. Tekalp and Y. Altunbasak, “Closed-Form Connectivity-Preserving Solutions for Motion Compensation Using 2-D Meshes,” IEEE Trans. on Image Processing, Vol. 6, No. 9, pp. 1255-1269, September 1997.
[8]A. M. Tekalp, P. V. Beek, C. Toklu and B. Gunesl, “Two-Dimensional Mesh-Based Visual-Object Representation for Interactive Synthetic/Natural Digital Video,” Proceedings of the IEEE, Vol. 86, No. 6, pp. 1028-1051, June 1998.
[9]A. M. Tekatp, H. Gokcetekin and I. Cetasun, “2D Mesh-Based Detection and Representation of an Occluding Object for Object-Based Video,” Speech and Signal Processing Proceedings, IEEE International Conference, Vol. 4, pp. 1915-1918, 2000.
[10]A. M. Tekalp and Y. Altunbasak, “Occlusion-Adaptive, Content-Based Mesh Design and Forward Tracking,” IEEE Trans. on Image Processing, Vol. 6, pp. 1270-1280, September 1997.
[11]A. M. Tekalp, T. Erdem and Y. Fu, “Tracking Visible Boundary of Objects Using Occlusion Adaptive Motion Snake,” IEEE Trans. on Image Processing, Vol. 9, No. 12, pp. 610-624, 2000.
[12]Y. Wang and O. Lee, “Active Mesh-a Feature Seeking and Tracking Image Sequence Representation Scheme,” IEEE Trans. on Image Processing, Vol. 3, No. 5, pp. 610-624, September 1994.
[13]P. Hsu, K. J. R. Liu and T. Chen, “A Low Bit-Rate Video Codec Based on Two-Dimensional Mesh Motion Compensation with Adaptive Interpolation,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 11, No. 1, pp. 111-117, January 2001.
[14]A. Nosratinia, “New Kernels for Fast Mesh-Based Motion Estimation,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 11, No. 1, pp. 40-51, January 2001.
[15]C. Lin and C. Y. Hsu, “A New Motion Compensation Method for Image Sequence Coding Using Hierarchical Grid Interpolation,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 4, No. 1, pp. 42-52, February 1994.
[16]P. Gerken, “Object-Based Analysis-Synthesis Coding of Image Sequences at Very Low Bit Rates,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 4, No. 3, pp. 228-235, June 1994.
[17]A. M. Tekalp, B. Gunsel and P. V. Beek, “Moving Visual Representations of Video Objects for Content-Based Search and Browsing,” Image Processing Proceedings, IEEE International Conference, pp. 502-505, 1997.
[18]S. Valette, I. Magnin and R. Prost, “Active Mesh for Video Segmentation and Objects Tracking,” Image Processing Proceedings, IEEE International Conference, Vol. 2, pp. 77-80, 2001.
[19]A. M. Tekalp, M. H. Gokcetekin, M. D. Harmanci and I. Celasun, “Mesh Based Segmentation and Update for Object Based Video,” Image Processing Proceeding, IEEE International Conference, Vol. 1, pp. 343-346, 2000.
[20]Gonzalez and Woods, Digital Image Processing, Addison Wesley, 1993.
[21]M. Sonka, V. Hlavac and R. Boyle, Image Processing, Analysis and Machine Vision, P.W.S, 1999.
[22]G. Wolberg, Digital image Warping, IEEE Computer Society Press, 1994.
[23]M. D. Berg, M. V. Kreveld, M. Overmars and O. Schwarzkopf, Computational Geometry-Algorithms and Applications, Springer, 1997.
[24]B. W. Shyr, “The Generation and Tracking of Mesh Object in Image Sequence,” M. S. thesis, National Chiao Tung University, Republic of China, 1998.
[25]K. N. Ngan, T. Meier and D. Chai, Advanced Video Coding: Principle and Techniques, Elsevier, 1999.
[26]R. Li, B. Zeng and M. L. Liou, “A New Three-Step Search Algorithm for Fast Block Motion Estimation,” IEEE Trans. on Circuits and Systems for Video Tech., Vol. 4, pp. 438-443, August 1994.
[27]K. K. Ma and S. Zhu, “A New Diamond Search Algorithm for Fast Block-matching Motion Estimation,” IEEE Trans. on Image Processing, Vol. 9, No. 2, pp. 287-290, February 2000.
[28]J. O’Rourke, Computational Geometry in C, Cambridge University Press, 1996.
[29]J. K Kim, J. W. Chung, J. H. Lee and J. H. Moon, “A New Vertex-Based Binary Shape Coder for High Coding Efficiency,” Signal Processing: Image Communication, Vol. 15, pp. 665-684, 2000.
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top