Inverse Tone Mapping Curve Adaptation for HDR Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for inverse tone mapping struggle to accurately convert standard dynamic range (SDR) content to high dynamic range (HDR) without introducing artifacts, particularly in handling highlight pixels and maintaining image quality, as they often fail to distinguish between light sources and diffusive reflections.
Innovation Solution
The proposed method involves detecting highlights and specular reflections in a scene to adaptively modify the inverse tone mapping curve, using a virtual source display to determine control points for the tone mapping curve, and applying modified multi-scale processes to reduce noise and artifacts, ensuring accurate conversion while maintaining image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If existing inverse tone mapping methods are used to convert SDR to HDR, then dynamic range expansion is achieved, but artifacts and noise are introduced in highlight regions
Solution Approach 1:
The patent segments the image processing into multiple scales (coarse to fine) and separates highlight detection from general tone mapping. By processing different spatial frequencies at different stages, the method expands dynamic range while suppressing artifacts that would otherwise appear in highlight regions.
Solution Approach 2:
The patent applies local adaptation by detecting highlights in specific regions and adjusting the tone mapping curve locally rather than globally. This allows different parts of the image to have different mapping characteristics, preserving highlight quality while maintaining overall dynamic range expansion.
2Productivity
If simple inverse tone mapping is applied, then conversion speed is maintained, but accuracy in distinguishing light sources from reflections is poor
Solution Approach 1:
The patent performs preliminary highlight detection and scene classification before applying the full inverse tone mapping process. By pre-identifying highlight regions and estimating scene luminance characteristics, the method prepares necessary parameters in advance, enabling accurate processing without sacrificing real-time performance.
Solution Approach 2:
The patent employs dynamic adaptation by adjusting the tone mapping curve based on detected scene characteristics and highlight properties. The mapping parameters are not fixed but dynamically adjusted according to local image content, improving accuracy while maintaining computational efficiency through adaptive rather than exhaustive processing.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
Novel methods and systems for inverse tone mapping are disclosed. A scene can be analyzed to obtain highlight detection from bright light sources and specular reflections. An inverse tone mapping curve can be calculated based on the lower dynamic range and higher dynamic range displays. Multi-scale filtering can be applied to reduce noise or artifacts.