Adaptive Noise Reduction Using Edge Keeping Index
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems apply uniform noise processing to all pixels, leading to excessive obscuration of detailed areas with high noise processing and inadequate noise reduction in areas with simple details, affecting image quality.
Innovation Solution
An image processing apparatus and method that determine noise processing levels based on an edge keeping index (EKI), using an EKI generating unit to calculate edge intensity values and adjust luminance values accordingly, ensuring proper noise processing by differentiating between pixels with varying detail complexities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If high-level noise processing is applied uniformly to all pixels, then noise is effectively reduced in areas with simple details, but image details in areas with complicated details become excessively obscured
Solution Approach 1:
The patent applies different noise processing levels to different regions of the image based on local characteristics. It calculates an Edge Keeping Index (EKI) for each pixel to determine whether the pixel belongs to a simple or complicated detail area, and then applies appropriate noise processing strength accordingly. This resolves the contradiction by making the noise processing adaptive to local image characteristics rather than uniform across the entire image.
Solution Approach 2:
The patent dynamically adjusts the noise processing level for each pixel based on real-time calculation of edge intensity and EKI values. The processing strength is not fixed but varies dynamically according to the local complexity of image details, allowing the system to optimize between noise reduction and detail preservation on a per-pixel basis.
2Loss of information
If low-level noise processing is applied uniformly to all pixels, then image details are preserved, but noise is not effectively restrained in areas with simple details
Solution Approach 1:
The patent identifies regions with simple details (low EKI values) and applies stronger noise processing specifically to those regions while maintaining lower processing levels in complicated detail areas. This localized approach ensures that noise is effectively restrained where needed without sacrificing important image details elsewhere.
3Manufacturing precision
If different noise processing levels are applied to different pixels based on detail complexity, then image quality is improved, but processing complexity increases
Solution Approach 1:
The patent segments the image processing into distinct stages: calculating original EKI values, calculating first adjusted EKI values after initial noise processing, and using these to determine final processing levels. This segmentation allows the complex task of adaptive noise processing to be broken down into manageable steps that can be implemented efficiently.
Solution Approach 2:
The patent performs preliminary calculations of EKI values before the main noise processing step. By pre-calculating the edge keeping index and determining which pixels require stronger or weaker processing, the system prepares the necessary information in advance, making the actual noise processing more efficient and reducing overall computational complexity.
Data Source
AI summary
An image processing apparatus includes an edge keeping index (EKI) generating unit and a noise reducing unit. The image decoding unit decodes a data stream to generate a plurality of image comprising at least a current image having the target pixel. The adjusting unit, coupled to the image decoding unit, comprises an edge keeping index (EKI) generating unit, for generating an edge intensity value of the target pixel according to an original luminance value of the target pixel and an original luminance value of at least one neighboring pixel associated with the target pixel, and a noise reducing unit, coupled to the EKI generating unit, for determining a first adjusted luminance value of the target pixel according to the original luminance value of the target pixel and the original luminance value of the at least one neighboring pixel associated with the target pixel, and for generating a static adjusted luminance value of the target pixel according to the original luminance value, the first adjusted luminance value and a first adjustment value of the target pixel. The adjustment value is determined by the edge intensity value.