HDR Image Encoding via LDR Decomposition and Transfer Function Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in efficiently encoding and decoding high dynamic range (HDR) images due to their complex nature, which complicates compression and transmission compared to standard low dynamic range (LDR) images.
Innovation Solution
The method involves decomposing HDR images into a lower dynamic range (LDR) image and either an exposure map or a look-up table (LUT), encoding these components, and then reconstructing the HDR image at the receiver end using the decoded LDR image and recomposition information, such as an exposure map or LUT, to facilitate efficient compression and transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If HDR images are encoded directly, then image quality is preserved, but encoding complexity and computational requirements increase significantly
Solution Approach 1:
The HDR image encoding process is segmented into two distinct stages: first encoding the LDR version of the image using conventional compressors, then separately encoding the transfer function parameters. This segmentation allows each component to be processed independently, reducing the overall encoding complexity while preserving the ability to reconstruct high-quality HDR images.
Solution Approach 2:
The transfer function parameters that define the HDR characteristics are extracted from the full HDR image data and encoded separately. This extraction isolates the critical HDR information from the general image data, allowing the LDR image to be encoded using standard, less complex compressors while the HDR-specific parameters are handled through a separate, simpler encoding process.
2Ease of manufacture
If HDR images are compressed using conventional compressors, then encoding simplicity is maintained, but compression efficiency and image quality deteriorate
Solution Approach 1:
The encoding process is divided into two parts: LDR image encoding using conventional compressors (maintaining simplicity) and separate transfer function parameter encoding (improving efficiency). This segmentation allows the system to leverage existing simple encoders while adding a targeted approach for HDR optimization.
Solution Approach 2:
The method changes the parameter representation by separating the transfer function parameters from the image data. This parameter change allows for more efficient encoding of HDR-specific information while maintaining compatibility with conventional LDR encoding pipelines, thus improving overall compression efficiency without sacrificing encoding simplicity.
3Productivity
If HDR images are decomposed into LDR image and transfer function parameters, then encoding efficiency improves, but reconstruction accuracy may be compromised
Solution Approach 1:
The transfer function parameters are extracted and preserved as separate encoded data. This extraction ensures that the critical HDR reconstruction information is maintained independently, allowing for accurate reconstruction without being lost or degraded in the compression process.
Solution Approach 2:
The decoding process uses feedback by applying the decoded transfer function parameters to the decoded LDR image to reconstruct the HDR image. This feedback mechanism ensures that the reconstruction accuracy is maintained by using the same transformation parameters that were used during the original HDR to LDR conversion.
4Measurement precision
If full HDR data is transmitted, then image quality is maintained, but transmission bandwidth requirements increase
Solution Approach 1:
The method extracts and transmits only the essential transfer function parameters separately from the LDR image data. This extraction reduces the total data volume significantly compared to transmitting full HDR data, while still preserving the capability to reconstruct high-quality HDR images at the receiving end.
Solution Approach 2:
Instead of transmitting the full HDR image data, the system transmits a compressed LDR copy along with compact transfer function parameters. This copying approach maintains image quality through accurate reconstruction while dramatically reducing the bandwidth requirements by transmitting only the essential reconstruction information.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
Various implementations address high dynamic range ("HDR") images. In one particular implementation, a low dynamic range ("LDR") image is generated from an HDR image, and information is generated allowing the HDR image to be reconstructed from the LDR image. The LDR image and the information are encoded. In another implementation, a signal or signal structure includes an LDR section including the encoded LDR image, and an information section including the encoded information. In another implementation, the encoded LDR image and the encoded information are both decoded. The HDR image is then reconstructed based on the decoded LDR image and the decoded information.