HDR Image Encoding via Backlight and Tone-Mapped Residual Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for encoding high-dynamic-range (HDR) images are inefficient due to the need for multiple encoding schemes, which limits dynamic range and correlation with the original image, and are not backward compatible with traditional apparatuses.
Innovation Solution
The method involves encoding a backlight image and a tone-mapped residual image, where the residual image is obtained by dividing the HDR image by the backlight image, and then tone-mapping and scaling the residual image to achieve efficient encoding and backward compatibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a tone-mapping operator is applied to the input HDR image and traditional encoding schemes are used, then backward compatibility is achieved, but the dynamic range is limited and coding performance deteriorates
Solution Approach 1:
The HDR image is segmented into two components: a backlight image representing the luminance component and a residual image containing the remaining information. This segmentation allows the backlight image to be encoded with high precision to preserve dynamic range, while the residual image can be tone-mapped and encoded using traditional schemes for backward compatibility, thus resolving the contradiction between compatibility and coding performance.
Solution Approach 2:
The luminance component (backlight image) is extracted from the HDR image and encoded separately with high precision. This extraction allows the most critical dynamic range information to be preserved without limitation, while the remaining residual image can be processed with traditional tone-mapping methods, maintaining backward compatibility without sacrificing overall coding performance.
2Manufacturing precision
If the residual image is not tone-mapped, then high dynamic range is preserved, but the residual image cannot be displayed by traditional apparatus
Solution Approach 1:
Different parts of the image data are treated with different quality requirements: the backlight image (luminance component) is encoded with high precision to preserve dynamic range, while the residual image is tone-mapped to lower precision suitable for traditional display devices. This local differentiation of quality requirements allows both high dynamic range preservation and broad display compatibility.
3Adaptability or versatility
If traditional encoding schemes are used directly on HDR images, then backward compatibility is maintained, but the correlation with the original HDR image weakens
Solution Approach 1:
The encoding process uses an asymmetric approach where the backlight image and residual image are encoded with different methods and precisions. The backlight image maintains high correlation with the original HDR image through precise encoding, while the tone-mapped residual image provides backward compatibility. This asymmetric treatment preserves maximum image correlation while ensuring compatibility.
Data Source
AI summary
The present invention generally relates to a method and device for encoding an image. The method comprises: —encoding (12) a backlight image determined (11) from the image; —obtaining (13) a residual image by dividing the image by a decoded version of the backlight image, —obtaining (16) a tone-mapped residual image by tone-mapping the residual image; and —encoding (19) the tone-mapped residual image. The invention relates also to a method and device for decoding a bitstream representing a residual image calculated by dividing an image by a backlight image.