Virtual Image Artifact Correction via Disparity Map
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for synthesizing cyclopean virtual images from stereo images often produce artifacts such as streaks, blocks, and halos due to color mixing discrepancies between corresponding pixels, leading to inaccurately synthesized virtual images with incorrect transparency effects and aliasing.
Innovation Solution
A disparity map is generated from stereo images to detect artifacts, and individual projected images are determined based on this map. A difference map is computed to identify artifacts, and a target patch is created as a composite of background and foreground exemplar patches from the input images to replace the source patch and correct the detected artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing methods are used to synthesize cyclopean virtual images from stereo images, then virtual images can be generated from stereo cameras offset relative to the display window, but artifacts (streaks, blocks, halos) are introduced that diminish image quality
Solution Approach 1:
The patent segments the virtual image synthesis process into multiple stages: generating initial virtual images from stereo cameras, detecting artifacts through difference maps, and correcting artifacts by replacing source patches with target patches. This segmentation allows the system to maintain the ability to generate virtual images from offset cameras while systematically removing quality-degrading artifacts.
Solution Approach 2:
The patent applies local quality improvement by detecting artifacts at specific pixel locations using difference maps and correcting only those localized regions. Instead of processing the entire image uniformly, the system identifies artifact-prone areas through color mixing discrepancy detection and applies targeted corrections using exemplar patches from regions without artifacts, thereby improving overall image quality while preserving accurate regions.
2Adaptability or versatility
If color mixing from different background colors in stereo images is used to synthesize virtual images, then virtual images can be created with transparency effects, but incorrect transparency effects and aliasing are introduced due to color mixing discrepancies
Solution Approach 1:
The patent implements feedback by computing difference maps that compare corresponding pixels from left and right stereo images. This feedback mechanism identifies color mixing discrepancies that indicate artifacts, allowing the system to detect and correct inaccurate transparency effects and aliasing by replacing problematic pixels with corrected values derived from exemplar patches.
Solution Approach 2:
The patent changes the parameter representation by working with difference maps and exemplar patches rather than directly manipulating original pixel values. By transforming the problem into identifying and correcting parameter deviations (color mixing discrepancies), the system can maintain accurate transparency effects while eliminating artifacts caused by incorrect color mixing.
Data Source
AI summary
Artifacts are detected in a cyclopean virtual image generated from stereo images. A disparity map is generated from the stereo images. Individual projected images are determined based on the disparity map and the corresponding stereo images. A difference map is computed between the individual projected images to indicate the artifacts. A source patch in the virtual image is defined relative to an artifact. A replacement target patch is generated using a split-patch search technique as a composite of a background exemplar patch and a foreground exemplar patch. Each exemplar patch may be identified from an image patch selected from at least one of the stereo images. The source patch of the virtual image is replaced by the replacement target patch to correct the detected artifact.