Abstract
In this thesis, we propose a novel algorithm to refine depth maps and generate multi-view video sequences from two-view video sequences for modern autostereoscopic display. In order to generate realistic contents for virtual views, high-quality depth maps are very critical to the view synthesis results. Therefore, refining the depth maps is the main challenging problem in the task. We propose an iterative depth refinement algorithm, including error detection and error correction, to correct errors in depth map. The error types are classified into across-view color-depth-inconsistency errors and local color-depth-inconsistency errors. Then, we correct the error pixels based on sampling local candidates. Next, we apply a trilateral filter that considers intensity, spatial and temporal terms into the filter weighting to enhance the temporal and spatial consistencies across frames. So the virtual views can be synthesized according to the refined depth maps. To combine both warped images, disparity-based view interpolation is introduced to alleviate the translucent artifacts. Finally, a directional filter is applied to reduce the aliasing around the object boundaries. Finally, the high-quality virtual views between the two views are generated. We demonstrate the superior image quality of the synthesized virtual views by using the proposed algorithm over the state-of-the-art view synthesis methods through experiments on benchmarking image and video datasets.