Adaptive Reshaping for Video Signal Detail Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video codecs fail to preserve visual details when encoding video data at low bit depths, leading to loss of details in high and low luminance ranges, and the metadata required for high-quality decoding is often too large for downstream devices to process effectively.
Innovation Solution
Implementing adaptive reshaping techniques in video codecs that dynamically allocate codewords based on image content, using power functions and lookup tables to approximate reshaping functions, allowing for efficient encoding and decoding of video signals across varying dynamic ranges while minimizing metadata transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If video data is encoded at low bit depth, then device complexity and metadata size are reduced, but visual details in high and low luminance ranges are lost
Solution Approach 1:
The patent applies preliminary tone mapping and adaptive reshaping during the encoding stage to pre-adjust the luminance distribution and allocate codewords according to the human visual system's sensitivity. This preliminary action ensures that visual details are preserved before compression, allowing low bit depth encoding to maintain quality without requiring complex metadata for post-processing correction.
Solution Approach 2:
The patent dynamically changes encoding parameters including bit depth, color space transformation matrices, and transfer functions based on the content's luminance characteristics. By adapting these parameters to match the HVS sensitivity at different luminance levels, the system achieves efficient compression while preserving visual details, resolving the contradiction between low complexity and information loss.
2Loss of information
If adaptive reshaping with dynamic codeword allocation is implemented, then visual details are preserved across luminance levels, but device complexity and processing requirements increase
Solution Approach 1:
The patent segments the luminance range into multiple regions (dark, mid-tone, bright) and applies different codeword allocation strategies to each segment based on HVS sensitivity. This segmentation allows the system to preserve visual details where most needed while using simpler encoding in less sensitive regions, reducing overall processing complexity while maintaining quality.
Solution Approach 2:
The patent applies local quality enhancement by allocating higher precision codewords to luminance regions where the human visual system is most sensitive (particularly mid-tones and transitions) and using coarser quantization in less sensitive regions. This localized approach preserves visual details effectively while minimizing the overall processing burden on downstream devices.
3Ease of operation
If metadata size is reduced for downstream device processing, then ease of operation improves, but the ability to achieve high quality decoding is compromised
Solution Approach 1:
The patent creates a simplified copy of the essential transfer function characteristics that can be represented with minimal metadata. Instead of transmitting complete lookup tables or complex transformation matrices, the system encodes the key parameters that define the reshaping behavior, allowing downstream devices to reconstruct the necessary transformation with reduced metadata while maintaining decoding quality.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
Statistical values are computed based on received source images. An adaptive reshaping function is selected for one or more source images based on the one or more statistical values. A portion of source video content is adaptively reshaped, based on the selected adaptive reshaping function to generate a portion of reshaped video content. The portion of source video content is represented by the one or more source images. An approximation of an inverse of the selected adaptive reshaping function is generated. The reshaped video content and a set of adaptive reshaping parameters defining the approximation of the inverse of the selected adaptive reshaping function are encoded into a reshaped video signal. The reshaped video signal may be processed by a downstream recipient device to generate a version of reconstructed source images, for example, for rendering with a display device.