3D Lookup Table Generation from Local Image Patches
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems struggle with generating tone-mapping information for raw image data captured by mobile devices, often resulting in under-exposed and over-exposed regions, as well as the creation of halos and other artifacts, due to the image-dependent nature of tone-mapping processes.
Innovation Solution
A 3D lookup table is generated by identifying image and gain map patches centered around anchor points, creating intensity-gain curves, and combining them to provide tailored tone-mapping adjustments, minimizing under-exposure, over-exposure, and artifact creation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional tone-mapping processes are used on raw image data, then the image can be converted to displayable formats, but under-exposed and over-exposed regions occur along with halos and artifacts
Solution Approach 1:
The patent divides the image into multiple local patches and processes each patch independently to generate local tone-mapping curves. This segmentation approach allows the system to adapt to local lighting conditions and avoid the global tone-mapping problems that cause under-exposure, over-exposure, halos, and artifacts in traditional methods.
Solution Approach 2:
The patent applies different tone-mapping characteristics to different regions of the image by generating local curves for each patch. This local quality approach ensures that each region is processed according to its specific luminance characteristics, preventing the uniform processing that leads to exposure problems and artifacts in traditional global tone-mapping methods.
2Loss of information
If raw image data with wider dynamic range is processed, then more detail is preserved, but the conversion to displayable formats creates exposure problems
Solution Approach 1:
The patent dynamically adjusts the tone-mapping parameters for each local patch based on its specific luminance distribution. By generating adaptive local curves rather than applying fixed global parameters, the system preserves detail from the wide dynamic range while correctly exposing each region for display conversion.
Solution Approach 2:
The patent changes the tone-mapping parameters locally for each patch based on its luminance characteristics. This parameter adaptation allows the system to maintain the rich detail information from the wide dynamic range raw data while transforming it into properly exposed displayable formats without the exposure problems of traditional methods.
3Device complexity
If global tone-mapping curves are generated, then processing is simpler, but image-specific characteristics are lost and artifacts appear
Solution Approach 1:
The patent segments the image into patches and generates local tone-mapping curves for each, replacing the single global curve approach. While this increases processing complexity, it eliminates the artifacts and loss of image-specific characteristics that occur with global methods by adapting to local image characteristics.
Data Source
AI summary
A method includes obtaining an image and a gain map associated with the image. The method also includes identifying image patches in the image and corresponding gain map patches in the gain map. Different image patches are centered around different anchor points in the image. The method further includes, for each image patch and its corresponding gain map patch, generating an intensity-gain curve for the associated anchor point. The intensity-gain curve specifies (i) gain values based on the corresponding gain map patch for intensity values up to a threshold intensity value and (ii) gain values based on one or more input parameters for intensity values above the threshold intensity value. In addition, the method includes combining the intensity-gain curves to generate a 3D lookup table, which identifies the gain values for the anchor points in the image at each of multiple intensity values.