A novel depth spatial temporal consistency enhancement algorithm for high compre

In free viewpoint video system based on multiview<br />video plus depth, inconsistency with depth video need to be<br />eliminated to ensure high-quality virtual view generation and<br />compression performance. The preprocessing method proposed<br />can compensate both spatial and temporal depth information<br />inaccuracy by using Bayesian probability model and Rival<br />penalized competitive learning in Self-Organizing Maps. Firstly,<br />each gray v