Image Resolution Adjustment with Symmetric Pixel Replication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image scaling techniques often result in image degradation, especially when upscaled, as they alter original pixel data, which is undesirable in applications like medical imaging where accuracy is critical, and complicate smoothness and calculation.
Innovation Solution
An image resolution adjustment method that selects one of four scaling modes based on source and display resolutions, using specific decision rules to generate scaling value sets for horizontal or vertical scaling, ensuring replicated pixels or lines are arranged symmetrically and smoothly, maintaining the original content's integrity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Shape
If interpolation is used to upscale images, then smooth edges are achieved, but original pixel data changes resulting in image degradation for accuracy
Solution Approach 1:
The patent segments the image scaling process into two distinct phases: a scaling phase that replicates pixels and lines to achieve the target resolution, and a smoothing phase that selectively applies smoothing only to specific regions. This segmentation allows the system to maintain original pixel data integrity in critical areas while still achieving smooth edges where appropriate.
Solution Approach 2:
The patent applies local quality by implementing region-based smoothing where different parts of the image receive different treatments. The smoothing operation is applied selectively based on local image characteristics and requirements, rather than uniformly across the entire image. This ensures that original pixel data is preserved in regions where accuracy is critical while achieving smooth edges in regions where it enhances quality.
2Measurement precision
If replication is used to upscale images, then original pixel data is maintained, but sawtooth edges appear resulting in degraded image quality
Solution Approach 1:
The patent divides the image processing into distinct scaling and smoothing phases, allowing replication to maintain pixel data integrity first, then applying smoothing selectively in the second phase to eliminate sawtooth edges only where necessary.
Solution Approach 2:
The patent applies smoothing partially rather than excessively - only to specific regions where edge smoothness is needed while leaving other regions unchanged. This partial application of smoothing maintains the benefits of replication (pixel data integrity) while mitigating the drawbacks (sawtooth edges) in a controlled manner.
3Manufacturing precision
If conventional scaling methods are used, then resolution conversion is achieved, but calculation complexity and smoothness control become troublesome
Solution Approach 1:
The patent simplifies the overall system complexity by segmenting the processing into two clear phases with distinct objectives. The scaling phase focuses solely on resolution conversion through pixel and line replication, while the smoothing phase handles quality enhancement. This segmentation makes the calculation logic more straightforward and easier to implement compared to conventional methods that attempt to handle both resolution conversion and smoothing in a single complex operation.
Data Source
AI summary
There is disclosed an image resolution adjustment method comprising the steps of: performing a predetermined selection rule to select one of scaling modes based on a source resolution Vi and a display resolution Vo to generate a scaling value set for outputting a scaled image; and determining whether a smoothing process to determine whether or not the scaled image is smoothed based on application requirements, wherein the scaling modes comprises a first scaling mode, a second scaling mode and a third scaling mode to generate each of the corresponding scaling value set, respectively. The scaling value set is provided to generate replicated pixels or lines for each pixel or line of the source image during a horizontal or vertical scaling period, and the replicated pixels or lines are arranged in horizontal or vertical symmetry in relation to its central region of the source image.