針對影片所做的三維梯度域階調映射與閃爍移除

將高動態範圍影像轉換成可以在一般螢幕播放的低動態範圍影像在近幾年已經是一個越來越重要的研究課題。雖然以二維梯度域為基礎的演算法具有將影像的細節和區域性對比保留的很完善的優點，如果我們將各自獨立處理的影像壓縮成為一個影片，在播放的時候，卻會發生大小不一的閃爍現象。造成這個現象的原因，主要是由於在做區域性轉換的時候，兩個具有一樣亮度值的小區域有可能會被轉換成為兩個不同亮度值。於是在影片播放的時候，這不同的亮度值，就使我們有閃爍的感覺。我們利用以三維梯度域為基礎的演算法，解決較小範圍的閃爍問題。接著再奠基於區域亮度的期望值和變異量這兩個影像的重要特點，提出一個以區塊為單位的亮度預測辦法，以解決大範圍的閃爍。整個運算過程在觀念上並不十分複雜，而我們處理過後的影片，不僅依然保留住豐富的細節資訊，同時也變的更加平順，幾乎徹底解決了閃爍的問題。

關鍵字

高動態範圍；二維梯度域演算法；閃爍問題

並列摘要

High dynamic range (HDR) radiance maps are more and more common and important nowadays. Using the 2D gradient-based algorithm to compress a HDR video sequence preserves fine details as well as local contrasts. However, the resulting LDR video shows small and large region flickers due to luminance variations from local tone mappings. We apply the 3D gradient compression to reduce the small region flickers. Then we propose the block-based luminance prediction scheme which manipulates both local intensity mean and variance of frames to remove unwanted artifacts due to large region flickers. The total computation course is simple conceptually and our final video sequence keeps plentiful visual contents along with smoothness without apparent flickers.

並列關鍵字

HDR ； 2D gradient algorithm ； flicker

參考文獻

[2] E. Reinhard, M. Stark, P. Shirley, and J. Ferwerda. “Photographic Tone Reproduction for Digital Images,” SIGGRAPH, pp. 267–276, 2002.

[3] E. Reinhard, G. Ward, S. Pattanaik, and P. Debevec. “High Dynamic Range Imaging,” Morgan Kanfmann

[4] R. Fattal , D. Lischinski, and M. Werman. “ Gradient Domain High Dynamic Range Compression,” ACM Transactions on Graphics, pp.249-256, 2002.

[5] H. Wang, R. Raskar and N. Ahuja. “Seamless Video Editing,” International Conference on Pattern Recognition (ICPR), pp. 858–861, 2004.

[6] K. K. Wang, A. Das and M. N. Chong. “Improved Flicker Removal through Motion Vectors Compensation,” International Conference on Image and Graphics (ICIG ), pp. 552–555, 2004.

國際替代計量

針對影片所做的三維梯度域階調映射與閃爍移除

全文下載

主題瀏覽