Methods and apparatus for constructing a 3D model of a scene and rendering new views of the scene

摘要

A three-dimensional model of a scene is reconstructed from a plurality of images of the scene taken from different viewpoints. A set of voxels that include the scene is defined, and each of the voxels in the set of voxels is processed. Each voxel is projected onto a set of pixels in each of the images. A first color-standard-deviation of not-occluded pixels is determined across all images. A second color-standard-deviation of not-occluded pixels is determined for each individual image. A mean of the second color-standard-deviations is determined across all images. A coloring threshold that is a function of the mean is established. If the first color-standard-deviation is less than the coloring threshold, the voxel is colored. Otherwise, the voxel is left transparent. The coloring threshold is established adaptively and depends on the color variation across the projection of the voxel in each individual image.