HDR Image Encoding via Luminance Scaling Metadata

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current HDR encoding technologies face challenges in efficiently encoding high dynamic range images for both HDR and LDR displays, requiring significant bits or complex two-layer approaches, and struggle to balance data amount, computational complexity, and artistic flexibility while maintaining compatibility with legacy systems.

Innovation Solution

A method of encoding HDR images by converting them into a lower luminance dynamic range, applying normalization, gamma conversion, and tone mapping, with metadata encoding the conversion functions to allow reconstruction of the original HDR image, enabling efficient encoding and decoding for various display brightness levels.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If two-layer HDR encoding (LDR image + illumination boost image) is used, then HDR image quality is improved, but device complexity increases

Engineering Contradiction:

ImproveHDR image qualityVSAvoidencoding complexity

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential HDR information (luminance scaling factors) from the full two-layer encoding approach. Instead of encoding complete LDR and illumination boost images, it separates and transmits only the critical metadata needed for HDR reconstruction, significantly reducing complexity while maintaining quality.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The encoding process is segmented into distinct stages: LDR encoding using existing standards, separate HDR parameter extraction, and independent HDR reconstruction. This segmentation allows legacy LDR infrastructure to be reused while adding minimal HDR-specific processing.

Inventive Principle:

Principle #1Segmentation

2Manufacturing precision

If more bits are used to encode HDR brightness above LDR range, then HDR rendering quality is improved, but data transmission requirements increase

Engineering Contradiction:

ImproveHDR brightness encoding precisionVSAvoiddata amount

Core Design Contradiction:

Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent changes the encoding parameter from direct high-precision luminance values to compact luminance scaling factors. By encoding relative brightness relationships rather than absolute HDR values, it achieves high HDR precision while using the same bit depth as legacy LDR systems.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

Instead of encoding HDR images directly with more bits, the patent inverts the approach by encoding LDR images with standard bits and storing HDR information as compact transformation parameters that can be applied during reconstruction, achieving high precision without increased data volume.

Inventive Principle:

Principle #13The other way round (Inversion)

3Manufacturing precision

If HDR encoding is designed for HDR displays, then HDR rendering quality is improved, but compatibility with legacy LDR displays decreases

Engineering Contradiction:

ImproveHDR rendering qualityVSAvoiddisplay compatibility

Core Design Contradiction:

Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The patent creates a universal encoding system where the same bitstream can be decoded for both LDR and HDR displays. The LDR decoder ignores HDR parameters and produces standard LDR output, while HDR decoders use the parameters to reconstruct high-dynamic-range images, achieving multi-display compatibility.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The LDR encoding is performed first using existing standardized processes, and HDR parameters are subsequently extracted and attached to the same bitstream. This preliminary LDR encoding ensures legacy compatibility while the added HDR parameters enable enhanced functionality for capable displays.

Inventive Principle:

Principle #10Preliminary action

4Ease of operation

If existing LDR decoders are used without modification, then ease of operation is maintained, but HDR image reconstruction capability is lost

Engineering Contradiction:

Improvedecoder usabilityVSAvoidHDR information retrieval

Core Design Contradiction:

Ease of operationVSLoss of information

Solution Approach 1:

The patent creates a copy of the LDR decoding path that preserves exact legacy behavior, while adding a parallel HDR reconstruction path. Existing LDR decoders continue to operate unchanged on the copied LDR stream, maintaining ease of operation, while HDR-capable systems can access the original HDR parameters for high-quality reconstruction.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS11949921B2Methods and apparatuses for encoding an HDR images, and methods and apparatuses for use of such encoded images

Publication Date: 2024.04.02 KONINKLIJKE PHILIPS NV

AI summary

Methods and apparatuses to encode both high dynamic range images and low dynamic range images. The input video may also convert the high dynamic range image to an image of lower luminance dynamic range by applying either a scaling the high dynamic range image to a predetermined scale of the luma axis or by applying a sensitivity tone mapping which changes the brightnesses of pixel colors falling within a subrange or by applying a gamma function or by applying an arbitrary monotonically increasing function mapping the lumas.