Methods and Systems for Volumetric Reconstruction Based on a Confidence Field

An exemplary volumetric reconstruction system accesses captured color and depth data for a surface of an object in a real-world capture space. The captured color and depth data is captured by capture devices positioned with respect to the real-world capture space to have different vantage points of...

Full description

Saved in:
Bibliographic Details
Main Author Castaneda, Oliver S
Format Patent
LanguageEnglish
Published 29.08.2019
Subjects
CALCULATING
COMPUTING
COUNTING
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:An exemplary volumetric reconstruction system accesses captured color and depth data for a surface of an object in a real-world capture space. The captured color and depth data is captured by capture devices positioned with respect to the real-world capture space to have different vantage points of the surface of the object. Based on the captured color and depth data, the volumetric reconstruction system generates reconstructed color and depth data for a volumetric reconstruction of the surface of the object. The generating of the reconstructed color and depth data includes allocating voxel nodes for surface points on the object within a voxel data store, determining and storing a confidence field value for each voxel node that accounts for a distance factor and a noise-reduction factor, and determining the reconstructed color and depth data using a raytracing technique based on the stored confidence field values within the voxel data store.
Bibliography:Application Number: US201815906795