HDR Image Encoding with Dynamic Range Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital image formats are not optimized for human vision capabilities, particularly in terms of luminance range and color gamut, leading to inefficient encoding and decoding of high dynamic range (HDR) images, which results in suboptimal image quality and increased data requirements.
Innovation Solution
An image encoder and decoder system that dynamically maps floating-point image values to integer values based on the extreme values of each pixel's luminance, using a floating point-to-integer converter and encoder, and an integer-to-floating point converter, respectively, to adapt the range of integer values to the distribution of floating-point values, thereby improving resolution and reducing data requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If floating-point values are mapped to integer values using a static mapping method, then the encoding process is simple, but the image quality is suboptimal and data requirements increase
Solution Approach 1:
The patent applies dynamic mapping by adjusting the mapping parameters based on the actual distribution of floating-point values in the image. The encoder determines the minimum and maximum floating-point values and uses these to dynamically calculate the mapping parameters, allowing the system to adapt to different image content rather than using a fixed static mapping approach.
Solution Approach 2:
The patent changes the mapping parameters based on the statistical properties of the image data. By calculating the minimum and maximum values of the floating-point image data and using these to determine the mapping parameters, the system optimizes the transformation from floating-point to integer representation according to the specific characteristics of each image.
2Measurement precision
If a wide range of integer values is used to accommodate HDR luminance range, then the luminance range is covered, but the resolution for narrow distributions is reduced
Solution Approach 1:
The patent dynamically adjusts the mapping parameters based on the actual minimum and maximum floating-point values in the image. This allows the system to use a narrower range of integer values when the floating-point values are concentrated in a small range, thereby preserving resolution, while still accommodating the full HDR luminance range when necessary.
Solution Approach 2:
The system transitions from a static mapping approach to a dynamic one where the mapping range is adjusted according to the actual data distribution. This dynamic adaptation allows the encoder to optimize the balance between range coverage and resolution based on the specific characteristics of each image.
3Adaptability or versatility
If floating-point format is used to store HDR images, then the dynamic range is accommodated, but the data size increases
Solution Approach 1:
The patent transforms the data representation by changing from floating-point format to integer format with dynamically determined mapping parameters. This parameter change allows the system to maintain the ability to represent HDR dynamic range while significantly reducing the data size, as integer values require fewer bits than floating-point values.
Solution Approach 2:
The patent uses a simpler integer data representation instead of complex floating-point format. This substitution reduces the computational complexity and storage requirements while maintaining the essential functionality of representing HDR images through the dynamic mapping approach.
4Adaptability or versatility
If standard digital image formats are used, then compatibility is maintained, but human vision capabilities are not optimized
Solution Approach 1:
The patent applies a logarithmic mapping transformation that better matches the non-linear perception of human vision. By changing the mapping function from linear to logarithmic, the system optimizes the representation of luminance values according to how humans actually perceive brightness differences, improving image quality for HDR content.
Data Source
AI summary
An image encoder includes an extreme value determiner, a floating point-to-integer converter and an encoder. The extreme value determiner determines minimal and maximal values of a floating point image value of each pixel of a part of an image, an image or a group of images. The floating point-to-integer converter maps the floating point image value of each pixel to an integer image value. The minimal floating point image value is mapped to a minimal integer image value of a predefined range of integer image values and the maximal floating point image value is mapped to a maximal integer image value of the predefined range of integer image values. The encoder encodes the integer image value of each pixel to obtain and provide encoded image data of the part of the image, the image or the group of images.