Variable Density Transparency Matte for Image Compositing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for compositing a presenter with presentation material in long distance learning often obscure or degrade the readability of text and other details, particularly when using chroma key techniques, and fail to maintain essential contrast, leading to confusion in distinguishing between foreground and background elements.
Innovation Solution
The method converts the foreground image into a variable density transparency 'matte' using edge detection, where higher edginess indicates low transparency and smoother areas indicate high transparency, allowing the background to show through while maintaining color and contrast, and uses this matte to compute a weighted average for compositing with the background.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If chroma key or color range compositing is used to place graphics over background, then background detail is prevented from showing through solid graphics areas, but this causes smooth varying color areas to become completely opaque, preventing background visibility and integration
Solution Approach 1:
The patent applies different transparency properties to different regions of the foreground image based on local characteristics. Edge detection identifies regions with high spatial frequency (text, details) and assigns them opaque properties, while smooth gradient regions are assigned transparent properties. This local differentiation resolves the contradiction by making only the necessary areas opaque for readability while preserving background visibility in smooth areas.
Solution Approach 2:
The patent transforms the binary transparency parameter into a continuous parameter based on edge strength. Instead of uniform opacity or complete transparency, the transparency value varies continuously across the image, with opaque regions where edges are detected and transparent regions where gradients are smooth. This parameter transformation enables simultaneous achievement of text readability and background integration.
2Object-generated harmful factors
If simple blending or dissolve is used to composite two images, then background visibility is maintained, but confusion arises in distinguishing which detail belongs to which image
Solution Approach 1:
The patent applies spatially varying transparency based on local image characteristics. By detecting edges and assigning opaque properties specifically to edge regions, the patent creates clear visual separation between foreground and background only where needed, while maintaining transparency in smooth areas. This local quality approach eliminates blending confusion while preserving necessary background visibility.
3Reliability
If chroma key technique is used to place graphics over background, then solid graphics areas remain completely opaque, but this prevents background detail from showing through areas that should be transparent for better integration
Solution Approach 1:
The patent makes transparency a local property determined by edge detection rather than a global property. Smooth gradient areas are identified as suitable for transparency to enable background integration, while edge regions maintain opacity for readability. This local quality approach simultaneously achieves contrast preservation and adaptability for integration.
Solution Approach 2:
The patent introduces dynamic transparency adjustment based on the detected edge strength at each location. The transparency parameter is not fixed but adapts locally according to the spatial frequency content, enabling the system to automatically determine the appropriate balance between readability and integration for each region of the image.
Data Source
AI summary
The present invention converts an image into a transparency, or “foreground image”, on which the readability of text and other detail is preserved after compositing with a background, while maintaining color information of broad areas of the image.