Adaptive Image Compression via Photo-Realism Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image compression technologies fail to adaptively control compression based on the degree of photo-realism in images, leading to suboptimal compression rates that can result in artifacts or inefficient use of compression resources.
Innovation Solution
A method and system that compute a photo-realism value for an image based on pixel luminance distribution, calculating a configuration parameter to configure an encoder for adaptive compression, allowing different compression levels for various image tiles based on their photo-realism scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a high compression level is applied to all image areas, then compression efficiency is improved, but artifacts appear in photo-realistic regions reducing image quality
Solution Approach 1:
The patent applies different compression levels to different regions of the image based on their photo-realism characteristics. Photo-realistic regions are identified and assigned lower compression levels to preserve quality, while non-photo-realistic regions receive higher compression levels to maximize efficiency. This regional differentiation resolves the contradiction by allowing high compression overall while protecting critical photo-realistic areas from artifacts.
Solution Approach 2:
The image is segmented into multiple regions based on photo-realism analysis. The system divides the image into photo-realistic and non-photo-realistic segments, then applies distinct compression parameters to each segment. This segmentation enables selective compression strategies that balance overall efficiency with local quality preservation in photo-realistic areas.
2Manufacturing precision
If a low compression level is applied to preserve photo-realism, then image quality is maintained, but compression efficiency decreases and resources are wasted
Solution Approach 1:
The patent implements quality preservation selectively only in photo-realistic regions rather than uniformly across the entire image. By identifying and targeting only the photo-realistic portions for lower compression, the system maintains image quality where it matters most while applying more aggressive compression to non-photo-realistic areas, thus avoiding resource waste.
Solution Approach 2:
Instead of applying low compression uniformly (excessive action), the patent applies low compression only partially to photo-realistic regions. This partial action approach achieves sufficient quality preservation in critical areas while avoiding the inefficiency of applying the same conservative compression strategy to all image regions, thereby improving overall compression efficiency.
3Device complexity
If uniform compression parameters are used for the entire image, then device complexity is reduced, but adaptability to different image regions is lost
Solution Approach 1:
The patent performs preliminary photo-realism analysis and region classification before the compression process. By pre-identifying photo-realistic and non-photo-realistic regions and assigning compression parameters in advance, the system enables adaptive compression without adding significant complexity during the actual compression execution. This preliminary preparation resolves the contradiction between simplicity and adaptability.
Data Source
AI summary
A method of controlling a compression of an image according to a degree of photo-realism in the image, comprising computing a value indicating a degree of photo-realism for an input image based on a distribution of pixel luminance values within the image, calculating a configuration parameter based on the photo-realism value for the image, and configuring an encoder to compress the image according to the configuration parameter.