H.264中有兩個錯誤隱藏的做法─「權重像素值平均法」(WPVA)跟「以邊界契合為基礎的動作向量復原法」(BMBMVR) ─分別是用來隱藏在「內圖」(intra picture)跟「間圖」(inter picture)的錯誤「巨集區塊」(macroblock)。然而，「權重像素值平均法」會模糊掉邊緣以至於造成不好的視覺效果，而「以邊界契合為基礎的動作向量復原法」會因為「H.264巨集區塊還可以再分區域來做動作補償」的機制，使得它對一整個巨集區塊所復原那個動作向量往往不適合此巨集區塊中每個次巨集區塊。為了解決「權重像素值平均法」的問題，我們提出「單一邊緣保留法」(SER)，它會以影像分析的方式判別此錯誤的巨集區塊是否在「單一邊緣」(Single Edge)上，並且以「方向性內插法」(Directional Interpolation)來達到保留邊緣的效果。此外，為了解決「以邊界契合為基礎的動作向量復原法」的問題，我們提出「次巨集區塊動作向量復原法」(SMBWMVR)，它會復原錯誤的巨集區塊下每個次巨集區塊的動作向量。再者，我們還修改「方均錯誤最小化雙向量動作補償法」(MMSE DMC)來把它整合到H.264以加強錯誤復原的能力。「方均錯誤最小化雙向量動作補償法」雖然對於動態區域有著很好的抑制錯誤傳遞的效果，但是對於靜態區域則不然。為了改善它對於抑制靜態區域錯誤傳遞的效能，我們對它裡面的模式權重等式做修改。

　　Two error concealment schemes Weighted Pixel Value Averaging (WPVA) and Boundary Matching Based Motion Vector Recovery (BMBMVR), which are adopt by H.264/AVC, are used to conceal erroneous MacroBlocks (MBs) in intra pictures and inter pictures, respectively. However, WPVA would blur edges so as to cause bad visional effects, and BMBMVR just recovers a motion vector for an entire MB, causing that the recovered motion vector is usually unsuitable for each sub-MBs of the erroneous MB due to the region-based motion compensation mechanism in H.264. To deal with the problem of WPVA, we propose Single Edge Reserving (SER), which detects whether MBs are on Single Edge by using image-analysis methods, and reserves them by using Directional Interpolation. Besides, to deal with the problem of BMBMVR, we propose Sub-MacroBlock-Wise Motion Vector Recovery (SMBWMVR), which recovers lost motion vectors for each sub-MB of the erroneous MB. Further, in order to enhance H.264 error resilience, we modify the Minima Mean Square Error Double-vector Motion Compensation (MMSE DMC) algorithm to adapt it into H.264. MMSE DMC provides good performance in the reduction of error propagation for moving regions, but not for static regions. To improve its performance for static regions, we modify its mode-weighting equation.

[1] “Draft ITU-T Recommendation and Final Draft International Standard of Joint Video Specification (ITU-T Rec. H.264 | ISO/IEC 14496-10 AVC),” in Joint Video Team (JVT) of ISO/IEC MPEG and ITU-T VCEG, JVTG050r1, 2003.

[2] T. Wiegand, H. Schwarz, A. Joch, F. Kossentini, and G. J. Sullivan, “Rate-constrained coder control and comparison of video coding standards,” IEEE Trans. Circuits Syst. Video Technol., vol. 13, pp. 688–703, July 2003.

[3] S. Kumar, L. Xu, M. K. Mandal and S. Panchanathan, Overview of Error Resiliency Schemes in H.264/AVC Video Coding Standard, To appear in Elsevier Journal of Visual Communication and Image Representation (Special issue on H.264/AVC Video Coding Standard), 2005

[4] P. Salama, N. B. Shroff, and E. J. Delp, “Error concealment in encoded video,” in Image Recovery Techniques for Image Compression Applications. Norwell, MA: Kluwer, 1998.

[5] W. M. Lam, A. R. Reibman, and B. Liu, “Recovery of lost or erroneously received motion vectors,” in Proc. ICASSP, vol. 5, Mar. 1993, pp. 417–420.

[6] V. Varsa, M. M. Hannuksela, and Y.-K. Wang, “Non-Normative Error Concealment Algorithms,” ITU-T VCEG-N62, 2001.

[7] Y.-K.Wang, M. M. Hannuksela, V. Varsa, A. Hourunranta, and M. Gabbouj, “The error concealment feature in the H.26L test model,” in Proc. ICIP, vol. 2, Sept. 2002, pp. 729–732.

[8] Chang-Su Kim, JongWon Kim, Katsavounidis I., Kuo C.-C.J., ”Robust MMSE video decoding: theory and practical implementations,” IEEE Trans. Circuits Syst. Video Technol, vol. 15, No. 1, January 2005

[9] Till Halbach, “Performance comparison: H.26L intra coding vs. JPEG2000,” in
ISO/IEC JTC1/SC29/WG11 and ITU-T SG16 Q.6, JVT 4th Meeting Klagenfurt,
Austria, July 2002.

[10] C.-S. Kim, R.-C. Kim, and S.-U. Lee, “Robust transmission of video sequence using double-vector motion compensation,” IEEE Trans. Circuits Syst. Video Technol., vol. 11, no. 9, pp. 1011–1021, Sep. 2001.

[11] W. Zeng and B. Liu, “Geometric structured based error concealment with novel applications in block-based low-bit-rate coding,” IEEE Transactions on IEEE Trans. Cir. Syst. Video Technol., Vol. 9(4), pp. 648 -665, 1999.