Adaptive Filtering in View Synthesis Prediction for Multiview Video Coding

View synthesis prediction has been studied to achieve efficient inter-view prediction. Existing view synthesis prediction methods generate the predicted pictures by using pictures decoded at the other views and geometric information of the scene. However, it is difficult to obtain such geometric information correctly. In addition, these conventional methods have no ability to compensate the inter-view difference in image signals caused by individual camera characteristics and the nonLambert reflection of objects. The method proposed herein can compensate both the interview signal mismatch and incorrect depth information by using an asymmetrical adaptive filter and the weighted average of wiener filter and the median filter. The proposed compensation process is applied to a geometrically compensated picture to minimize the effect of warping-based view synthesis. Experiments show that the proposed method reduces the bitrate by up to 7% relative to view synthesis prediction based on the general adaptive filtering method.