Image Enhancement Using Adaptive High-Frequency Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low-quality images often suffer from blurred edges and reduced definition due to environmental factors or scaling, leading to a decrease in image quality, which existing image enhancement algorithms struggle to effectively address.
Innovation Solution
An image enhancement method that extracts high-frequency components, calculates an adaptive enhancement threshold based on these components, and applies enhancement to pixels with values greater than or equal to the average pixel value, thereby improving image details and overall definition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional image enhancement algorithms are used to adjust brightness, contrast, saturation, and hue, then overall image quality is improved, but edge details and local definition remain blurred
Solution Approach 1:
The patent applies local quality enhancement by computing adaptive thresholds separately for different image regions based on local histogram analysis. Different enhancement parameters are applied to different regions (edge regions vs. non-edge regions), allowing edge details to be enhanced with higher contrast while maintaining appropriate enhancement for other areas, thus resolving the contradiction between overall quality improvement and local detail preservation
Solution Approach 2:
The image is segmented into different regions (edge regions and non-edge regions) based on gradient magnitude or local variance. This segmentation allows the enhancement algorithm to treat different regions differently, applying more aggressive enhancement to edge regions while using milder enhancement for other areas, thereby improving both overall quality and local definition simultaneously
2Manufacturing precision
If enhancement is applied to all pixels uniformly, then overall image brightness and contrast improve, but noise is amplified and natural image characteristics are lost
Solution Approach 1:
The patent computes local enhancement thresholds based on regional statistics rather than applying a global threshold. This allows the algorithm to adapt the enhancement strength to local characteristics, applying stronger enhancement only where needed (in regions with sufficient signal) and avoiding enhancement in noisy regions, thus improving definition without excessive noise amplification
Solution Approach 2:
The enhancement parameters (thresholds, scaling factors) are dynamically changed based on local image characteristics such as local variance, gradient magnitude, or histogram distribution. This adaptive parameter adjustment ensures that enhancement is applied appropriately across different regions, improving definition while minimizing noise amplification in areas where it would be harmful
Data Source
AI summary
An image enhancement method is provided. The method includes obtaining an image, extracting high-frequency components of the image, calculating an average pixel value of pixels corresponding to the extracted high-frequency components, and performing enhancement on pixels in the image that have pixel values greater than or equal to the calculated average pixel value to obtain an enhanced image. Apparatus and non-transitory computer-readable storage medium counterpart embodiments are also provided.