作者: Sergey Smirnov , Atanas Gotchev
DOI: 10.1117/12.2086895
关键词:
摘要: Depth Image-based Rendering (DIBR) is a popular view synthesis technique which utilizes the RGB+D image format, also referred to as view-plus-depth scene representation. Classical DIBR prone dis-occlusion artefacts, caused by lack of information in areas behind foreground objects, appear visible synthesized images. A number recently proposed compensation techniques have addressed problem hole filling. However, their computational complexity does not allow for real-time and may require additional user input. In this work, we propose hole-compensation technique, works fully automatically perceptually-correct manner. The applies two-layer model given imagery, specifically tailored rendering with free viewpoint selection. main two components are an adaptive layering depth into relative 'foreground' 'background' layers be rendered separately blending filtering aimed at creating function aliasing cancellation during process composition. implementation turns ordinary images true 3D representations, visualization y-around