3D Scene Modeling Using Depth Map Surfaces and Voxel Transparency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for modeling a real scene in three dimensions, such as Visual Hull and Space carving, face challenges with concave area modeling errors and photo-consistency estimation issues, especially in scenes with significant light variations, leading to inaccurate voxel elimination.
Innovation Solution
A method that generates surfaces from depth maps in world space, discretizes the bounding volume into voxels with transparency attributes, and modifies these attributes based on ray intersections to create a resulting surface, focusing on geometric criteria to improve modeling fidelity and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If Visual Hull technique is used to reconstruct 3D scene from multiple views, then a 3D model can be obtained through volumetric intersection of de-projected silhouettes, but modeling errors occur in concave areas
Solution Approach 1:
The patent replaces photometric-based methods (Visual Hull using silhouettes and Space Carving using photo-consistency) with a geometric-based method. Instead of using image intensity and color information, the invention uses depth maps and geometric ray intersections to determine voxel transparency, thereby avoiding modeling errors in concave areas that occur with photometric methods.
2Productivity
If Space carving method is used with photo-consistency estimation, then voxels can be eliminated to reconstruct 3D model, but erroneous eliminations occur in areas with significant light variations
Solution Approach 1:
The patent changes the fundamental parameter used for voxel classification from photometric parameters (image intensity, color consistency) to geometric parameters (depth map values, ray intersection positions). By using depth information and geometric relationships instead of light-based measurements, the method achieves accurate voxel elimination regardless of lighting conditions.
3Ease of operation
If threshold value is used to determine photo-consistency of voxels, then voxel classification can be performed, but modeling errors are induced by inappropriate threshold selection
Solution Approach 1:
The patent substitutes the threshold-based photometric classification mechanism with a geometric classification mechanism. Instead of comparing image intensities against a threshold, the method uses depth map geometry and ray intersection tests to determine whether voxels are inside or outside the modeled object, eliminating the need for threshold selection and associated errors.
Data Source
AI summary
The invention relates to a method for modeling a scene from a plurality of maps representative of depth, each map representative of depth being associated with a view of the said scene according to a particular viewpoint, To optimize the fidelity and precision of the scene modeling, the method comprises the following steps:generating, for each map representative of depth, a surface representative of the said depth map in world space;estimating a volume bounding all surfaces generated, the said bounding volume being discretised into a plurality of voxels;associating attributes representative of transparency with the said voxels of the bounding volume; andgenerating a resulting surface according to the attributes representative of transparency associated with the said voxels.The invention also relates to a device for processing the data representative of the corresponding depth.