臺灣博碩士論文加值系統

近年來，由於網際網路之蓬勃發展，使用網路的人口已經快速地普及，這促使網路的應用更加豐富且多元化，網路多媒體的應用也更顯重要，然而，像是視訊內容這樣龐大的資料量，若是不先透過壓縮處理是根本無法經由網路的傳輸來播放使用，有鑑於此，關於視訊壓縮的相關研究也正被熱烈的討論，相關標準也陸續地被制訂。在視訊內容上，由於連續的畫面間存在著相當大的相關性，這意味著連續的畫面間有很多的資訊是一直被重複的表示出來，因此若是能先行有效地剔除這些重複表現的資訊，便能大大地減少視訊內容的資料量，甚至加快後續壓縮編碼上處理的過程，這樣的概念便是所謂的『動態估測與移動補償之技術』，該作法已被廣泛的使用在許多視訊壓縮標準中，例如：MPEG-1、MPEG-2、H.261和H.263。
在視訊壓縮的議題中，若是使用動態估測的技術去壓縮視訊畫面間的影像資訊，最重要的要求是要能有效且快速的計算出移動補償資訊，如此才能進一步的契合現今的網路視訊應用，例如視訊會議系統或是隨選視訊系統。
現今所有動態估測的衡量標準均是已具有最小平方誤差(MSE)之方塊為估測點，而移動補償係針對估測誤差作JPEG壓縮，眾所周知的事實是JPEG係以DCT將空間域之相關性消除來達到資訊量減少之目的；所以所謂最佳估測點應是將估測誤差壓縮後資料量最小之點；但傳統以最小平方法為判斷依據之結果是否真的會有最小資料量呢？經初步研究顯示，以Full search估測法顯示超過50%的最佳點其壓縮後的資訊量均不是最少，且比真正最佳點之資訊量平均超過10%以上，由此可知影響壓縮後資訊量之因素應是估測誤差之相關係數而非MSE。因此我們試著去找出一個新的移動估測的判斷標準以便得到更好的移動補償。因此我們試著去找出移動補償之間的關連性當作移動估測的新判斷標準。

In recent years, the internet has been in widespread use and the number of internet subscribers increased quickly. Hence a lot of applications on the network have been developed, multimedia programs especially. Whereas the original video content always takes up considerable storage and transmission time which doesn’t suit for network application, many video compression standards have been drawn up in the literature
Due to the temporal redundancy of the video sequences, motion estimation / compensation has been widely used in many interframe video coding protocols to reduce the required bit rates for transmission and storage of video signals by eliminating it, such as the MPEG-1, MPEG-2, H.261 and H.263.
The performance and speed of the interframe motion estimation method for video sequence compression are the important issues especially in networking application such as video conference and video on demand.
Today all motion estimating method find out the estimating point which has minimal Mean Square Error, and motion compensation aim at estimating error to do JPEG. compression. As everyone knows, JPEG employs DCT to eliminate the correlation of spatial domain. So the best motion estimating point is the point which has the minimal compressed data size. In some alalyses show that over 50% best estimating point do not have the minimal compressed data size. So the factor which effects the compressed data size is correlation coefficient and not MSE. Hence, we try to define a new criterion for motion estimation which can get better motion compensation with less compressing bit rate. To reach this goal, we try to find out the correlation among the motion compensation as the new criterion for motion estimation.

Content………………………………………………………………………………i
List of Figures………………………………………………………………………iii
List of Tables…………………………………………………………………iv
Abstract……………………………………………………………………………v
1.Introduction………………………………………………………………………1
1.1 Motivation and Recent Related Research…………………………………1
1.2 Summary of the Thesis…………………………………………………2
1.3 Organization of the Thesis…………………………………………………3
2.Moving Picture Expert Group…………………………………………………5
2.1 Moving Picture Expert Group Overview………………………………5
2.2 Intra Frame Encoder Algorithm……………………………………………9
2.2.1 Color Space Transform…………………………………………… 10
2.2.2 Sampling……………………………………………………………12
2.2.3 Discrete Cosine Transform…………………………………………14
2.2.4 Quantization…………………………………………………………15
2.2.5 Entropy Coding…………………………………………………17
3.Motion Estimation………………………………………………………………23
3.1 Motion Estimation Technique…………………………………………23
3.2 Several Existing Search Algorithms for Block-Matching…………………27
3.2.1 Full Search Algorithm………………………………………………27
3.2.2 Uniform Search Algorithm…………………………………………29
3.2.3 Center-Biased Search Algorithms……………………………………30
3.2.4 Minima Bounded Area Search Algorithm……………………………39
4.Confusion and Consideration of Mean Square Error……………………………42
4.1 Compression Data Size V.S. Mean Square Error…………………………42
5.Correlative Difference And Mean Square Error…………………………………46
5.1 Motion Compensation With Correlative Difference………………………46
5.2 Correlative Difference And Minimal Compressed Size…………………48
5.3 Correlative Difference With Mean Square Error…………………………49
5.4 Simulation Result of Correlative Difference………………………………50
6.A Novel Motion Estimating Method With Correlative Difference……………53
6.1 The Algorithm of UCBDS-CDN…………………………………………55
6.2 Simulation Results of UCBDS-CDN Algorithm……………………58
7.Conclusion……………………………………………………………………60
Reference……………………………………………………………………………62
List of Figures
Figure 2-1 An example of the group of MPEG……………………………………6
Figure 2-2 The Flow Chart of MPEG Encoder……………………………………8
Figure 2-3 Flow chart of JPEG compression………………………………………9
Figure 2-4 YCbCr in (a)4:4:4 format (b)4:2:2 format (C)4:2:0format…………13
Figure 2-5 The flow chart of Entropy Coding………………………………………18 Figure 2-6 The order of Zig-Zag Scan………………………………………………20
Figure 3-1 Motion compensation encoding diagram………………………………24
Figure 3-2 Illustration of a motion displacement search for motion estimation.25
Figure 3-3 The scope of full search……………………………………………28
Figure 3-4 (a) the search window (w=7)
(b) 225 search points for the full search algorithm………………28
Figure 3-5 The three-step search……………………………………………………30
Figure 3-6 The new three-step search…………………………………………32
Figure 3-7 Search pattern of the FSS………………………………………………34
Figure 3-8 Two different search paths of FSS……………………………………35
Figure 3-9 UCBDS search pattern…………………………………………………37
Figure 3-10 Example of UCBDS…………………………………………………38
Figure 3-11 A search example using the center-biased MIBAS…………………41
Figure 4-1 Five different motion compensation value……………………………42
Figure 5-1 Two matrixes with different correlative difference…………………49
Figure 6-1 Avg. Search Points and MSE for five Video Sequences………………54

List of Tables
Table 4-1 The results for compressed data size in FSA……………………………45
Table 5-1 Improved Ratio of the compressed data size in some criteria…………51
Table 6-1 The results for Football and Susie sequence……………………………53
Table 6-2 Improved Ratio of compressed data size………………………………58
Table 6-3 Search Points of each motion estimation…………………………………59

Reference
[1]T. Koga, K. Iinurna, A. Hirano, Y. Iijima, and T. Ishiguro, “Motion-compensated interframe coding for video conferencing,” in Proc. NTC 81, pp. C9.6.1-9.6.5, New Orleans, LA, Nov./Dec. 1981.
[2]Reoxiang Li; Bing Zeng; Liou, M.L., ”A new three-step search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Technol., Vol. 4, pp. 438-442, Aug. 1994.
[3]L. M. Po and W. C. Ma, ”A novel four-step search algorithm for fast block motion estimation,” IEEE Trans. Circuits Stst. Video Tcehnol., vol. 6, pp. 313-317, June 1996.
[4]J. Y. Tham, S. Ranganath, M. Ranganath, and A. A. Kassim, “A novel unrestricted center-biased diamond search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Technol., vol. 8, pp. 369-377, Aug. 1998.
[5]Christopoulos, V.; Cornelis, J., “A center-biased adaptive search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Technol., vol. 10, pp. 423-426, Apr. 2000.
[6]W. Li and E. Salari, “Successive elimination algorithm for motion estimation,” IEEE Trans. Image Processing, vol. 4, pp. 105-107, Jan. 1995.
[7]X. Q. Gao, C. J. Duanmu, and C.R. Zou, “A multilevel successive elimination algorithm for block matching motion estimation,” Image Processing, vol. 9, pp. 501-504, Mar. 2000.
[8]I. Ismaeil, A. Docef, F. Kossentini, and R. Ward, “Efficient motion estimation using spatial and temporal motion vector prediction,” in Proc. ICIP 99, pp. 70-74, Oct. 1999.
[9]L. Luo, C. Zou, X. Gao, and Z. He, “A new prediction search algorithm for block motion estimation in video coding,” IEEE Trans. Consumer Electron., vol. 43, pp. 56-61, Feb. 1997.
[10]J. M. Jou, P. Y. Chen, and J. M. Sun, “The gray prediction search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Technol., vol. 9, pp. 843-848, Sept. 1999.
[11]C. Cafforio and F. Rocca, “Methods for measuring small displacements of television images,” IEEE Trans. Inform. Theory, vol. IT-22, no.5, pp. 573-579, Sept. 1976.
[12]ISO/IEC 11172-2: Coding of Moving Pictures and Associated Audio for Digital Storage Media at up to About 1.5 Mbits/s — Part 2: Video, International Standards Organization, August 93.
[13]Joan L. Mitchell, William B. Pennebaker, Chad E. Fogg, and Didier J. LeGall, MPEG Video Compression Standard, International Thomson Publishing.
[14]ISO/IEC DIS 13818-2: Generic Coding of Moving Pictures and Associated Audio Information — Part 2: Video, International Standards Organization, May 94.
[15]CCITT SGXV, “Description of reference model 8 (RM8),” Document 525, Working Party XV/4, Specialists Group on Coding for Visual Telephony, June 1989.
[16]Video Coding for Low Bit Rate Communication, ITU-T Rec. H.263, January 1998.
[17]Karel Rijkse, KPN Research, “H.263:Video Coding for Low-Bit-Rate Communication,” IEEE Communication Magazine, pp. 42-45, December 1996.
[18]ISO/IEC 10918-1: Information Technology — Digital Compression an dCoding of Contious-Tone Still Images: Requirements and Guidelines, International Standards Orgamization, Feb. 94.
[19]In, J.; Shirani, S.; Kossentini, F, “On RD optimized progressive image coding using JPEG”, Image Processing, IEEE Transactions on , Volume: 8 Issue: 11 , Nov. 1999, pp. 1630 —1638.
[20]Konstantinides, K.; Bhaskaran, V.; Beretta, G, “ Image sharpening in the JPEG domain”, Image Processing, IEEE Transactions on , Volume: 8 Issue: 6 , June 1999, pp. 874 —878.
[21]Lee, Y.L.; Kim, H.C.; Park, H.W, “Blocking effect reduction of JPEG images by signal adaptive filtering”, Image Processing, IEEE Transactions on , Volume: 7 Issue: 2 , Feb. 1998, pp. 229 —234.
[22]Yi-Huang Han; Jin-Jang Leou , “Detection and correction of transmission errors in JPEG images”, Circuits and Systems for Video Technology, IEEE Transactions on , Volume: 8 Issue: 2 , April 1998, pp. 221 —231.
[23]Sheng-Li Wei ,“CockTail Search (CTS) for Video Motion Estimation”, June 2001.