Novel Temporal and Spatial Error Concealment with Game Theory

通訊網路上封包遺失是無法避免的，在視訊傳送時如果有資料方塊遺失，解碼端一般藉由時間域的錯誤隱藏技術來回恢復它們，然而時間域的錯誤隱藏可能會產方塊效應的嚴重問題，因此為了更好的視覺品質，我們需要將這些影像除去它們的方塊效應。在本論文，我們提出八方向鑽石搜尋法來對錯誤方塊在時間域上進行恢復，接著我們讓帶有方塊效應的影像透過我們所提出的幾何形變方法來除去方塊效應。我們在系統架構下不斷地進行迭代處理，直到錯誤區塊被良好的恢復。此外，由於在解碼端很可能沒有原始的視訊影片，我們基於賽局理論提出一個可以讓恢復的視訊自身評估恢復品質的方法，透過這個賽局，我們可以離線式的找到納許均衡的策略，而事實上，排除耗時的暴力搜尋，從實驗中得出的納許均衡正是我們所提出的方法組合。之後，我們可以在需要執行錯誤隱藏時總是使用此納許均衡的解來進行恢復。

關鍵字

error concealment ； temporal ； spatial ； game theory

並列摘要

Packet loss is inevitable over communication network. When there are missing blocks in video transmission, the decoder usually recovers them by temporal error concealment technique. However, it might produce a serious problem of blocking effect in the concealed images. Therefore, these images need to be deblocked for better visual quality. In the thesis, we propose Eight-way Diamond Search to recover the error blocks in temporal. And then we let the recovered images go through a proposed deblocking method , Geometric Deformation. We iteratively process the error blocks under our system framework until they are well-recovered. Moreover, since we might not get the original video sequence in the decoder, we design a way to measure the performance of the recovered video with game theory. Through the game, we can find the Nash Equilibrium strategy profile offline. In fact, except the full search, Nash Equilibrium strategy profile is right our proposed methods. Afterwards, we can always do the error concealment by the Nash Equilibrium.

並列關鍵字

錯誤隱藏；時間域；空間域；賽局理論

參考文獻

[1] W. Kim, J. Koo, and J. Jeong, “Fine Directional Interpolation for Spatial Error Concealment,” IEEE Trans. Consumer Electronics, vol. 52, pp. 1050-1056, Aug. 2006.

[2] A. Sirikam, and W. Kumwilaisak, “New Spatial Error Concealment using Dynamic Texture Estimation and Geometric Interpolation,” IEEE Int. Conf. Multimedia and Expo, pp. 120-123, Jul. 2007.

[3] G. Zhai, J. Cai, W. Lin, X. Yang, and W. Zhang, “Image Error Concealment via Block-based Bilateral Filtering,” IEEE Int. Conf. Multimedia and Expo, pp: 621-624, Jun. 2008.

[4] P. Perez, M. Gangnet, and A. Blake, “Poisson Image Editing,” Proc. ACM SIGGRAPH, pp: 313-318, Jul. 2003.

[5] Z. Wang, Y. Yu, and D. Zhang, “Best neighborhood matching: an information loss restoration technique for block-based image coding systems,” IEEE Trans. Image Processing, vol. 7, pp: 1056-1061, Jul. 1998.

國際替代計量

Novel Temporal and Spatial Error Concealment with Game Theory

全文下載

主題瀏覽