Multiview Texturing for 3D Models Using Pose Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for texturing three-dimensional models using polygon meshes often result in inconsistencies and unnatural appearances due to projection misalignments and blending issues between source images, leading to blurred or misaligned textures.
Innovation Solution
A computer-implemented method that aligns source images using an approximate pose model with a global shift and grid-based warping function to minimize projection misalignments, determines a weighted average texel value from aligned pixels, and removes outlier pixels to generate a robust and consistent texture for polygon meshes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If source images are directly projected onto the polygon mesh without alignment, then the process is simple and fast, but projection misalignments occur causing blurred or misaligned textures
Solution Approach 1:
The patent applies preliminary action by performing image alignment before texture mapping. The approximate pose model is used to pre-align source images to the polygon mesh, correcting projection misalignments in advance. This ensures that when textures are mapped, the images are already positioned correctly, avoiding blurred or misaligned results without requiring complex real-time adjustment processes.
Solution Approach 2:
The patent utilizes parameter changes by employing an approximate pose model that transforms image coordinates to mesh coordinates through parameter transformations. The global shift and grid-based warping function adjust image parameters (position, orientation, scale) to match the polygon mesh geometry, enabling precise texture alignment while maintaining computational efficiency.
2Reliability
If all source image pixels are used in the weighted average, then the texture is comprehensive, but outlier pixels degrade texture quality
Solution Approach 1:
The patent applies the extraction principle by identifying and removing outlier pixels from the set of source image pixels before computing the weighted average. The outlier detection mechanism extracts and eliminates pixels that do not conform to the expected texture pattern, such as pixels from different materials, shadows, or artifacts. This ensures that only reliable pixels contribute to the final texture, maintaining quality consistency while processing a manageable number of pixels.
Solution Approach 2:
The patent implements feedback by using the weighted average result to validate and refine the outlier detection process. The feedback mechanism compares the generated texture against the source images and adjusts the outlier identification criteria accordingly. This iterative feedback loop ensures that outlier pixels are accurately identified and removed, while maintaining comprehensive texture information from valid source pixels.
Data Source
AI summary
Systems and methods for generating textures to be rendered in conjunction with a polygon mesh are provided. More particularly, a polygon mesh modeling a geographic area can be accessed. A plurality of source images depicting the geographic area can then be identified. The plurality of source images can be aligned to reduce projection misalignments between source images when the source images are projected to the polygon mesh. A texture can then be determined based at least in part on a weighted average of pixels in the plurality of source images corresponding to a point on the surface of the polygon mesh. Determining a texture can include removing at least one outlier pixel associated with a moving object from the weighted average.