HDR to LDR Tone Mapping Using Naka-Rushton and Laplacian Pyramid
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing tone-mapping methods for converting high dynamic range (HDR) images to low dynamic range (LDR) images face challenges in maintaining natural exposure and contrast, often resulting in artifacts like halos and unnatural representations, especially when dealing with varying lighting conditions.
Innovation Solution
The method involves applying a smoothing filter, a Naka-Rushton function, response mapping, and an inverse Naka-Rushton function to HDR images, utilizing parameters like semi-saturation constants and sensitivity controls to adapt to human vision characteristics, and employing Local Laplacian Filters or other edge-aware filters to preserve image details and reduce artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If a global tone mapping operator based on HVS is applied to HDR images, then the image can be displayed with realistic appearance, but local contrast is lost
Solution Approach 1:
The image is divided into multiple layers using Laplacian pyramid decomposition, allowing different processing to be applied to different frequency components. This segmentation enables preservation of local contrast in high-frequency layers while maintaining realistic appearance in low-frequency layers.
Solution Approach 2:
Different tone mapping operations are applied to different layers of the Laplacian pyramid. The low-frequency base layer undergoes global tone mapping for realistic appearance, while high-frequency detail layers preserve local contrast through selective processing and adaptive blending.
2Loss of information
If a local tone mapping operator based on bilateral filter is applied, then local image contrast is preserved, but artifacts like halos are generated
Solution Approach 1:
The bilateral filter operation is applied selectively to specific layers in the Laplacian pyramid decomposition rather than to the entire image. This segmentation approach preserves local contrast where needed while avoiding halo artifacts in other regions.
Solution Approach 2:
The parameters of the bilateral filter (such as spatial and range standard deviations) are adaptively adjusted based on local image characteristics and layer position in the pyramid. This dynamic parameter adjustment optimizes contrast preservation while minimizing artifact generation.
3Loss of information
If Laplacian pyramid-based tone mapping is applied, then image contrast is enhanced and shadows/highlights are boosted, but overall image exposure cannot be controlled
Solution Approach 1:
The processing framework integrates multiple functions: global tone mapping for exposure control, local contrast enhancement through Laplacian pyramid processing, and adaptive blending. This multi-functional system allows simultaneous control of both overall exposure and local contrast characteristics.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present application relates to a method for converting a high dynamic range, HDR, image file into a low dynamic range, LDR, image file, wherein the method comprises the following steps: - providing (110) an HDR image file to a processing unit (14); - applying (120), by the processing unit (14), a smoothing filter on the provided HDR image file; - applying (130), by the processing unit (14), a Naka Rushton function on the HDR image file using the filtered image file as eye adaptation intensity; - conducting (140), by the processing unit (14), a response mapping on the output of the Naka-Rushton function; and - applying (150), by the processing unit (14), an inverse Naka-Rushton function on the output of the response mapping.