Image Multiplexing Layout for Object Area Quality Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image multiplexing techniques often result in deterioration of image quality, particularly in areas of interest within an image where information is embedded, as users tend to focus attention on these areas, leading to conspicuous quality degradation.
Innovation Solution
An image processing apparatus that recognizes objects within an image and lays out multiplexed information in areas outside the object area, using a recognition unit and layout unit to select and embed information in non-object areas, thereby reducing image quality deterioration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If multiplexed information is embedded in an image using conventional techniques, then information can be multiplexed, but image quality deteriorates particularly in object areas
Solution Approach 1:
The patent applies local quality by differentiating between object areas and non-object areas in the image. Multiplexed information is selectively embedded only in non-object areas, while object areas maintain their original quality without any modifications. This localized approach ensures that the most visually important regions remain pristine while still achieving the goal of information embedding in other parts of the image.
Solution Approach 2:
The patent segments the image into two distinct regions: object areas and non-object areas. This segmentation is achieved through object recognition technology that identifies and separates the main subject from the background. By dividing the image in this manner, the system can apply different processing strategies to each region, embedding information only where it will not interfere with the visual quality of the primary subject.
2Quantity of substance
If multiplexed information is embedded in the entire image, then information capacity is maximized, but object area quality deteriorates
Solution Approach 1:
The patent implements local quality by restricting information embedding to non-object areas only. This selective approach prioritizes the quality of object areas over maximum information capacity. The system accepts a trade-off where information capacity is reduced compared to embedding throughout the entire image, but this ensures that object areas maintain their original visual quality without degradation.
3Loss of information
If changes are made to embed multiplexed information, then information can be multiplexed, but visual recognition of the image deteriorates
Solution Approach 1:
The patent applies local quality by concentrating all modifications exclusively in non-object areas. Since these regions are less critical for visual recognition, the harmful effects of embedding operations (such as pixel modifications and pattern additions) are confined to areas where they are least noticeable. This protects the visual recognition quality of the primary subject while still enabling information embedding.
Solution Approach 2:
The patent converts the potential harm of embedding operations into a benefit by strategically locating these operations in non-object areas. The modifications that would normally degrade image quality are redirected to background or less important regions, where they become imperceptible. This transforms what would be a harmful effect into a beneficial strategy for maintaining overall image quality while embedding information.
Data Source
AI summary
Reduce image quality deterioration of an object area that is gazed viewed by a user in a case where information is embedded in an image. An image processing apparatus including a recognition unit configured to recognize an object within an image and a layout unit configured to lay out information relating to an object recognized by the recognition unit in an area except for an object area for the recognized object within the image.


