High Dynamic Range Image Processing Using Region-Specific Noise Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High dynamic range imaging (HDRI) faces challenges in preventing image overlap and noise occurrence, especially under low illumination or high sensitivity conditions, which deteriorate image quality.
Innovation Solution
An image processing apparatus and method that classify low dynamic range images into saturation, under-saturation, and non-saturation regions, generate feature maps to identify moving and non-moving pixels, and apply different noise reduction filters to remove noise, while using a Debevec and Malik method to generate a radiance map with adjusted weights for moving and non-moving regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If multiple low dynamic range images with different exposure levels are captured to generate HDRI, then the luminance range representation is improved, but image overlap occurs due to subject movement or background changes
Solution Approach 1:
The patent segments the image into moving regions and non-moving regions by comparing feature maps generated from non-saturation areas of multiple LDR images. This segmentation allows differential processing where non-moving regions are used for reliable HDRI composition while moving regions are identified and handled separately to prevent misalignment artifacts.
Solution Approach 2:
The patent applies local quality by using different noise reduction filters for moving and non-moving regions. Non-moving regions undergo more aggressive noise reduction since they are reliable for HDRI generation, while moving regions receive different processing to preserve motion details and avoid overlap artifacts.
2Illumination intensity
If images are captured in low illumination or high sensitivity conditions to achieve proper exposure, then the exposure quality is improved, but noise occurs during HDRI generation which deteriorates image quality
Solution Approach 1:
The patent applies local quality by using different noise reduction filters for moving and non-moving regions. Non-moving regions undergo more aggressive noise reduction since they are reliable for HDRI generation, while moving regions receive different processing to preserve motion details and avoid overlap artifacts.
Solution Approach 2:
The patent performs preliminary noise reduction on individual LDR images before HDRI composition, and also performs noise reduction on the final HDRI radiance map. This multi-stage noise reduction approach addresses noise at different points in the processing pipeline to prevent its propagation and accumulation.
3Area of stationary object
If saturation regions are included in HDRI generation, then complete image coverage is achieved, but false detection of moving regions occurs reducing processing accuracy
Solution Approach 1:
The patent extracts and excludes saturation regions from the feature map generation process. By comparing only non-saturation regions to create feature maps, the system avoids the false detection problem that would occur in saturation areas, while still maintaining complete image coverage through separate handling of saturated pixels.
Data Source
AI summary
An image processing apparatus and an image processing method, the method including: providing one or more low dynamic range images having different exposure levels; classifying each of the one of more low dynamic range images as a saturation region, an under-saturation region, and a non-saturation region with respect to an intensity value; generating feature maps indicating a similarity between patterns of the non-saturation regions; generating final feature maps converting pixels included in the feature maps into moving pixels and non-moving pixels; defining moving regions and non-moving regions included in the one or more low dynamic range images; removing noise from the moving regions and the non-moving regions by using a first noise reduction filter for the moving regions and a second noise reduction filter for the non-moving regions; and generating a radiance map corresponding to the one or more low dynamic range images.