Scalable Video Encoding Adaptive Filtering and Upsampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video codecs face inefficiencies in encoding and decoding processes, particularly when dealing with high sample depth and color fidelity video content, as they often require reducing image fidelity to lower bit rates, leading to artifacts and suboptimal compression efficiency.
Innovation Solution
The implementation of adaptive lowpass filtering, sample depth upsampling, chroma upsampling, and remapping techniques in scalable video encoding and decoding tools to improve the efficiency of inter-layer residual video compression by adjusting sample depth, chroma sampling rates, and dynamic range, thereby reducing artifacts and enhancing compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If video is compressed to lower bit rate by removing detail, then bit rate is reduced, but video quality and fidelity deteriorate
Solution Approach 1:
The video signal is segmented into base layer and enhancement layer components. The base layer contains essential video information at lower quality, while the enhancement layer contains additional detail information. This segmentation allows selective transmission where the base layer can be transmitted at lower bit rates while the enhancement layer provides quality improvement when bandwidth permits.
Solution Approach 2:
The patent applies parameter changes by transforming video data between different sample depths (e.g., 8-bit to 10-bit) and chroma sampling rates (e.g., 4:2:0 to 4:2:2). These parameter transformations enable the video to be encoded at different quality levels while maintaining compatibility across multiple transmission scenarios.
2Loss of energy
If chroma sampling rate is reduced to fewer samples, then bit rate requirement decreases, but color fidelity deteriorates
Solution Approach 1:
Chroma information is segmented and treated separately from luma information. The chroma channels are downsampled independently to reduce bit rate while the luma channel maintains higher resolution. This selective segmentation allows the system to reduce overall bit rate while preserving perceived video quality, since the human eye is more sensitive to luma variations than chroma variations.
Solution Approach 2:
The patent employs parameter changes by converting between different chroma sampling rates (4:4:4, 4:2:2, 4:2:0) based on transmission requirements. The video can be encoded with full chroma resolution when quality is paramount, or with reduced chroma resolution when bit rate constraints exist, providing flexible adaptation to different application scenarios.
3Manufacturing precision
If sample depth is increased to represent values more precisely, then video quality and color gamut improve, but bit rate increases
Solution Approach 1:
The video encoding is segmented into base layer with lower sample depth (e.g., 8-bit) and enhancement layer with higher sample depth (e.g., 10-bit or 12-bit). The base layer provides acceptable quality at lower bit rates, while the enhancement layer adds precision for applications requiring higher quality. This segmentation allows receivers to discard the enhancement layer when bandwidth is limited, achieving efficient adaptive streaming.
Solution Approach 2:
The patent applies parameter changes by transforming video data between different sample depths through upsampling and downsampling operations. The video can be encoded at 10-bit or 12-bit precision when quality is the priority, or converted to 8-bit for more efficient compression. These parameter transformations enable flexible adaptation to different transmission scenarios and device capabilities.
4Manufacturing precision
If video is encoded with high sample depth and color fidelity, then video quality improves, but compression efficiency deteriorates
Solution Approach 1:
The video is segmented into base layer and enhancement layer, where the base layer is compressed efficiently at lower quality and the enhancement layer contains the additional fidelity information. This segmentation allows the majority of the video data to be compressed efficiently in the base layer, while only the differential enhancement information requires higher precision encoding, thereby improving overall compression efficiency compared to encoding the entire video at high fidelity.
Solution Approach 2:
The patent employs parameter changes by transforming video parameters (sample depth, chroma sampling rate) to match the requirements of different compression stages. The base layer uses parameters optimized for compression efficiency (lower sample depth, reduced chroma sampling), while the enhancement layer uses parameters optimized for quality (higher sample depth, full chroma sampling). This parameter adaptation enables the system to achieve both good compression efficiency and high video fidelity.
Data Source
AI summary
Techniques and tools for conversion operations between modules in a scalable video encoding tool or scalable video decoding tool are described. For example, given reconstructed base layer video in a low resolution format (e.g., 4:2:0 video with 8 bits per sample) an encoding tool and decoding tool adaptively filter the reconstructed base layer video and upsample its sample values to a higher sample depth (e.g., 10 bits per sample). The tools also adaptively scale chroma samples to a higher chroma sampling rate (e.g., 4:2:2). The adaptive filtering and chroma scaling help reduce energy in inter-layer residual video by making the reconstructed base layer video closer to input video, which typically makes compression of the inter-layer residual video more efficient. The encoding tool also remaps sample values of the inter-layer residual video to adjust dynamic range before encoding, and the decoding tool performs inverse remapping after decoding.