Multiplane Image Alpha and Texture Correction for Rendering Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multiplane images generated using existing methods often exhibit artifacts such as 'dark pixels' and 'black holes' when rendered on displays with different characteristics from the reference display, which degrade image quality.
Innovation Solution
A method involving alpha-channel normalization, texture-channel refinement, and texture-channel scaling is applied to multiplane images to adjust weights, alpha and texture values, ensuring that rendered images match the source image quality by minimizing differences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiplane images are rendered on displays with different characteristics from the reference display, then the images can be viewed on various displays, but artifacts such as 'dark pixels' and 'black holes' appear that degrade image quality
Solution Approach 1:
The patent applies parameter changes by adjusting alpha channel values and texture channel values in the multiplane image layers. Specifically, it normalizes alpha values to correct transparency-related artifacts and scales texture values to compensate for display characteristic differences, thereby eliminating dark pixels and black holes while maintaining image quality across different displays
Solution Approach 2:
The patent employs feedback mechanisms by comparing rendered image characteristics against reference display characteristics and iteratively adjusting the multiplane image parameters. The system uses weight optimization and value normalization based on the observed artifacts, creating a closed-loop process that continuously improves image quality across different display conditions
2Ease of manufacture
If existing multiplane image methods are used, then the rendering process is simple, but artifacts appear that degrade image quality
Solution Approach 1:
The patent applies preliminary action by pre-processing the multiplane image data before rendering. It performs alpha channel normalization and texture channel refinement in advance, adjusting weights and values to prevent artifact formation during the rendering process itself, thereby maintaining rendering simplicity while improving image quality
Solution Approach 2:
The patent modifies key parameters including alpha values, texture values, and layer weights before rendering. By normalizing alpha channels and scaling texture channels in the source multiplane image, it prevents artifact generation during rendering without complicating the rendering process
Data Source
AI summary
Image-processing technique directed at improving the quality of viewable images generated by rendering a multiplane image having a plurality of pixels and represented by a plurality of layers corresponding to different respective distances from the reference camera position. In an example embodiment, the image-processing technique includes one or more of the following operations: (A) for a first set of pixels, scaling respective weights of the layers to cause a sum of the scaled weights to be normalized to one; (B) for a second set of pixels, replacing respective alpha and texture values in the layers by the corresponding local average values; and (C) for a third set of pixels, scaling corresponding texture values in the layers such that, for the resulting viewable image rendered for the reference camera position, texture values of the third set match the respective texture values of the source image captured from the reference camera position.