Image Texture Uniformization for Multiview Object Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for combining multiple images to create textures for 3D objects often result in non-uniform color and brightness, leading to an uneven appearance when applied to 3D models, despite being dimensionally correct.
Innovation Solution
A method involving the determination of brightness values for visible points on an object across multiple images, generation of tile grids, calculation of brightness offset values for tile pairs, and application of offset matrices to normalize brightness across images, followed by transformation and reconstruction to create a uniformly textured 3D model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional image combination techniques are used to create textures from multiple images, then the texture can be generated with correct dimensions, but the color and brightness become non-uniform across the texture
Solution Approach 1:
The patent divides the texture into multiple tiles corresponding to different images, and processes each tile separately by calculating brightness offsets relative to a reference tile. This segmentation allows independent brightness normalization of each image region while maintaining overall texture coherence.
Solution Approach 2:
The patent introduces brightness offset values as adjustable parameters that are calculated for each tile pair. These offsets are then applied to normalize brightness across all tiles, transforming the original non-uniform brightness distribution into a uniform one while preserving dimensional accuracy.
2Area of stationary object
If multiple images are combined to create a texture, then coverage of the 3D object is improved, but adjacent regions show different color shades and contrast levels
Solution Approach 1:
The patent establishes a reference tile that serves as a brightness baseline, and calculates brightness offsets for all other tiles relative to this reference. This creates an equipotential brightness surface across the entire texture, ensuring consistent color appearance across all regions regardless of their source image.
3Productivity
If image stitching is performed without brightness normalization, then the processing time is reduced, but the resulting texture appears non-uniform when applied to 3D models
Solution Approach 1:
The patent performs brightness offset calculation and image transformation operations before final texture stitching. By preparing and normalizing brightness in advance for each tile, the system ensures uniform appearance in the final texture without requiring additional post-processing steps, thus maintaining efficiency while improving quality.
Data Source
AI summary
An image processing method is disclosed, comprising: receiving a plurality of images of an object; and, for each image, determining a brightness value for each point on the object that is visible, and generating a tile grid for the image. Each point that is visible in an image corresponds to one tile in the tile grid. For each tile pair where at least one point is visible in both tiles of the tile pair, a brightness offset value is determined. For each image, an offset matrix is generated. Values of the offset matrix comprise brightness offsets to apply to each tile of a tile grid for the image to normalize the brightness of the image with the other images. A plurality of reconstructed images is generated based on performing, for each image, transformation operations on the given image based on the corresponding offset matrix to generate a reconstructed image.