Multiview stereo refers to the task of reconstructing a 3D shape from calibrated overlapping images captured from different viewpoints. Various representations of 3D shape can be used. For example, dense 3D point cloud or surface mesh representations are common in applications that synthesize a new photorealistic image of the scene using computer graphic rendering techniques. The topics of multiview stereo and multi-baseline stereo matching share key concepts related to the recovery of dense 2D pixel correspondences in multiple images.
CITATION STYLE
Sinha, S. N. (2014). Multiview Stereo. In Computer Vision (pp. 516–522). Springer US. https://doi.org/10.1007/978-0-387-31439-6_203
Mendeley helps you to discover research relevant for your work.