- 學位論文
多視角視訊彩色影像錯誤隱藏演算法
Multiview video error concealment algorithm for color images
摘要
本文提出一種對於全畫面錯誤隱藏方法的改進。本方法是在有深度視訊資訊的多視角視訊。本文包括三個部分的修改。首先，錯誤隱藏中一對一的像素由從另一視訊中的像素做補償。視訊之間的光照差是由估計協調的運動向量資訊，和一個反向DIBR （Depth image based rendering）程序。這些程序是對於一個有深度視訊資訊的多視角視訊，針對一個最先進的全畫面遺失的錯誤隱藏方法所作的假設，提出更好的假設。其次，多對一個像素的產生是利用它們的深度信息資訊從而得到效能的改善。第三，找到空洞的像素所使用的估計的運動向量，即相鄰的像素及其可用的運動向量和其到目標空洞像素距離的加權函數。實驗結果表明，相對於最先進的方法，上述提出的方法組合比較好，而且提高了性能最大至6.08 dB。
並列摘要
An improved whole frame error concealment method for multiview video with depth is designed in the proposed thesis. The thesis includes three modification parts. First, the erroneous 1-to-1 pixels are concealed by the pixels from another view. The light differences between views are taken care by the information of motion vector of the projected coordination, and a reverse DIBR (Depth image based rendering) procedure. These procedures are based on assumptions that are better than the assumptions made in a state-of-the-art whole frame loss method for multiview video with depth. Second, the generation of the many-to-1 pixels is improved with their depth information. Third, the hole pixels are found using the estimated motion vectors, that is a weighted function of the neighboring available motion vectors and their distance to the target hole pixel. The experiment results show that compared to the state-of-the-art method, the combined system of three proposed methods is better, and improve the performance by 6.08 dB at maximum.