Digital Image Perceptual Detail Enhancement via Random Luminance Blending
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital images and videos often suffer from poor quality due to encoding artifacts like macro-block chunks, flat spots, and gradient banding, especially when upscaled or highly compressed, leading to a compromised viewer experience.
Innovation Solution
The technique involves adding strategically-placed artificial detail to digital images and videos, which are perceived as actual detail by the human eye, reducing the visibility of encoding artifacts by modifying pixel luminance values and blending them with random values to create a more natural and pleasing image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If bicubic or bilinear upscaling is used to fill in missing pixels at higher resolution, then the image can be displayed at greater resolution, but the smoothing effect causes significant blurriness
Solution Approach 1:
The patent modifies pixel luminance values by adding random variations within controlled ranges, transforming the smooth interpolated values from upscaling into values with perceptual detail. This parameter change converts the blurriness caused by upscaling into apparent sharpness without actually increasing the underlying resolution data.
Solution Approach 2:
The patent introduces an intermediary processing step between upscaling and display that adds artificial detail through random luminance modifications. This intermediary layer masks the blurriness introduced by upscaling, allowing the system to benefit from both high display resolution and perceived image sharpness.
2Manufacturing precision
If contrast is increased or images are sharpened to compensate for blurriness from upscaling, then perceived sharpness improves, but encoding artifacts are brought into sharper focus and video degradation is exacerbated
Solution Approach 1:
The patent applies different treatments to different regions by modifying luminance values locally at each pixel position with random variations. This local quality approach adds detail where needed without uniformly sharpening the entire image, thereby avoiding the amplification of encoding artifacts that occurs with global sharpening operations.
Solution Approach 2:
Instead of applying traditional sharpening filters that amplify high-frequency content and expose artifacts, the patent changes the luminance parameter of individual pixels by small random amounts. This subtle parameter modification creates perceived detail without the harsh side effects of conventional sharpening methods.
3Loss of substance
If encoding compression is applied to reduce file size, then storage and transmission efficiency improves, but visual detail and acutance decrease due to artifacts like macro-block chunks and posterization
Solution Approach 1:
The patent converts the harm caused by compression artifacts into a benefit by adding random luminance variations that mask the artificial patterns created by compression. The same computational approach that creates compression inefficiency is used to generate perceptual detail that hides the damage, turning a negative effect into a positive one.
Solution Approach 2:
The patent modifies the luminance parameter of compressed image pixels by adding random variations within controlled ranges. This parameter change restores perceptual detail and acutance that was lost during compression, without requiring re-compression or increasing the fundamental compression ratio.
Data Source
AI summary
A method for increasing the perceived quality of a digital image. The method includes receiving a first luminance value associated with a pixel located at a first pixel location in the digital image; generating a second luminance value based on a random number; blending the first luminance value and the second luminance value to generate an output luminance value; and displaying on a display device at the first pixel location or storing in a memory an output pixel having a brightness based on the output luminance value.