3D Surface Display Using Weighted Texel Blending
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately display objects from vantage points different from the original capture points, leading to artifacts and inaccuracies due to varying angles and occlusions in panoramic images.
Innovation Solution
A system that generates a 3D model of object surfaces from multiple vantage points, using ray tracing to project visual characteristics onto polygons, and applies weight values based on orientation and quality values to blend texels from different source images, addressing occlusions and artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If panoramic images are captured from multiple fixed vantage points, then complete surface coverage is achieved, but display accuracy deteriorates when viewing from uncaptured angles due to artifacts and occlusions
Solution Approach 1:
The system dynamically synthesizes images for any requested vantage point by blending texels from multiple source images based on calculated quality values, rather than being limited to fixed captured angles. This allows continuous adaptation to user-defined viewing positions while maintaining visual accuracy through real-time computation of blend weights based on surface orientation and capture geometry.
Solution Approach 2:
A 3D model serves as an intermediary structure that stores texels from multiple source images and enables synthesis of intermediate views. The model acts as a mediator between the discrete source images and the continuous space of possible viewing angles, allowing accurate reconstruction of surfaces from uncaptured perspectives through weighted combination of available texels.
2Adaptability or versatility
If texels from multiple source images are blended to fill uncaptured regions, then view coverage is improved, but visual accuracy deteriorates due to artifacts from inappropriate texel selection
Solution Approach 1:
The system applies different quality values to different texels based on their local properties and relationship to the requested vantage point. Each texel is evaluated individually using metrics such as surface orientation alignment, capture angle, and visibility, allowing the blending process to selectively weight texels based on their local suitability rather than applying uniform blending rules.
Solution Approach 2:
The system changes the weighting parameters of texels dynamically based on the requested vantage point and surface geometry. Quality values are computed as functions of multiple parameters including the angle between surface normal and view direction, the capture position relative to the surface, and the requested viewing position, allowing optimal texel selection that adapts to each specific synthesis scenario.
3Device complexity
If simple texel blending is used to generate views from uncaptured positions, then computational complexity is reduced, but display quality deteriorates due to artifacts and inaccuracies
Solution Approach 1:
The system performs preliminary organization of texels into a 3D model structure with associated quality metrics before image synthesis. By pre-computing and storing the spatial relationships, surface orientations, and quality values for each texel in the model, the system avoids repeated complex calculations during rendering, enabling high-quality synthesis with optimized computational flow.
Data Source
AI summary
A system and method is provided for displaying surfaces of an object from a vantage point different from the vantage point from which imagery of the object was captured. In some aspects, imagery may be generated for display by combining visual characteristics from multiple source images and applying greater weight to the visual characteristics of some of the source images relative to the other source images. The weight may be based on the orientation of the surface relative to the location from which the image was captured and the location from which the object will be displayed.