Stereo Vision for View Synthesis
نویسنده
چکیده
We propose a new method for view synthesis from real images using stereo vision The method does not explicitly model scene geometry and enables fast and exact generation of synthetic views We also re evaluate the requirements on stereo algorithms for the application of view synthesis and discuss ways of deal ing with partially occluded regions of unknown depth and with completely occluded regions of unknown tex ture Our experiments demonstrate that it is possible to e ciently synthesize realistic new views even from inaccurate and incomplete depth information
منابع مشابه
3D Scene Reconstruction by Stereo Methods for Analysis and Visualization of Sports Scenes
The 3D reconstruction of image and video scenes by stereo analysis is an important topic in computer vision research. In this talk, we first present some principles of stereo algorithms and recent developments. We then demonstrate two applications of stereo reconstruction for the analysis and visualization of human movement: (a) We employ depth maps derived from sport scenes for novel view synt...
متن کاملResearch Activities in 3D Content Generation
This report describes work on 3D content generation performed under the Austrian 3D-ConTourNet participation project Hyperion3D. We review key processing steps such as disparity computation, inter-camera mapping and virtual view synthesis, and give information on a publicly available database for evaluating 3D visual discomfort. Keywords—trifocal camera; 3D film; stereo matching; view synthesis...
متن کاملAdaptive Image-space Stereo View Synthesis
Stereo vision is becoming increasingly popular in feature films, visualization and interactive applications such as computer games. However, computation costs are doubled when rendering an individual image for each eye. In this work, we propose to only render a single image, together with a depth buffer and use image-based techniques to generate two individual images for the left and right eye....
متن کاملMulti-H: Efficient recovery of tangent planes in stereo images
The input of Multi-H are point correspondences with local affine transformations and the epipolar geometry. We use Matching On Demand with view Synthesis (MODS) method [3] since it provides accurate local affinities, the fundamental matrix F and point correspondences consistent with F. Using Homography from Affine transformation and Fundamental matrix (HAF) method [1] a homography is estimated ...
متن کاملImage-Based View Rendering for 3D Visual Communications
We present an image-based view rendering from multiple sample views. It is capable of explicitly handling uncovered area problems in view synthesis and an improved version of our previous work [8] toward quality enhancement with faster rendering. For high quality of image, we introduce a postprocessing scheme in stereo matching and reduce the total number of image resampling to one in rendering...
متن کامل