Contrast Correction Using Gradient-Weighted Statistics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing contrast enhancement techniques often produce disturbing artifacts near edges of large imaged objects, as they either fail to adequately expand weaker edges or introduce noise due to the selection of a scale parameter, and conventional methods do not effectively account for edge-preserving filtering in contrast correction.
Innovation Solution
An image processing apparatus that uses a pixel value mapper controlled by statistical data from a statistics collector, which weights contributions from surrounding pixel locations based on gradient detection, employing edge-preserving low pass filters to reduce the impact of pixel locations across edges, thereby reducing artifacts and enhancing image features within regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If contrast enhancement is applied to expand intensity variations, then edge visibility is improved, but artifacts appear near edges of large objects
Solution Approach 1:
The patent applies different processing treatments to different regions of the image based on local characteristics. By computing histograms locally and detecting edges, the system applies contrast enhancement only where needed while preserving edge regions, thus improving edge visibility without creating artifacts in homogeneous areas.
Solution Approach 2:
The image is segmented into different regions based on edge detection results. The patent divides the image into edge regions and non-edge regions, applying different histogram equalization strategies to each segment, which prevents artifacts from propagating across edge boundaries.
2Device complexity
If a fixed scale parameter is used for histogram computation, then processing is simplified, but optimal contrast enhancement cannot be achieved in all regions
Solution Approach 1:
The patent introduces dynamic adaptation by computing histograms at multiple scales and selecting the appropriate scale based on local image characteristics. Instead of using a fixed scale parameter, the system dynamically adjusts the histogram computation scale to match local features, achieving optimal contrast enhancement adaptively.
Data Source
Figure 1~2
Figure 1a
Figure 3~4
AI summary
An image processing apparatus collects local image statistics to determine parameters for contrast correction. The statistics for a pixel location is collected from pixel values for surrounding pixel locations. The weight assigned to contributions from surrounding is reduced dependent on image gradients between the pixel location and the surrounding pixel locations, less weight being given in the case of larger gradients. Pixel values are mapped under control of statistical data. A spread in the statistical data is used to determine the stretching factor of the contrast correction. Edge preserving filtering may be applied to squared deviations from the average pixel value in order to determine the spread with such weighted contributions for the determination of the stretching factor.