Selective Lossless Video Encoding for Text and Graphics Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face limitations in efficiently encoding compound videos that contain both text/graphics and natural-view content, particularly in maintaining high fidelity and compression efficiency, especially with the lack of a dedicated lossless coding mode in existing standards like HEVC.
Innovation Solution
The proposed solution involves a video encoding apparatus and method that selectively bypasses quantization and transform steps for text/graphics regions, using a lossless mode, while encoding natural-view content in a lossy mode, and includes a transform bypass and transform without quantization coding scheme to improve coding efficiency and fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a dedicated lossless coding mode is added to video coding standards, then the fidelity of text and graphics encoding is improved, but the device complexity and coding process complexity increase
Solution Approach 1:
The video frame is segmented into different regions (text/graphics regions and natural image regions), with different coding modes applied to each region. This allows lossless coding to be applied selectively only where needed, rather than to the entire frame, thereby improving text/graphics fidelity without unnecessarily increasing overall complexity.
Solution Approach 2:
The coding mode is made dynamic and adaptive, allowing the encoder to switch between lossless and lossy modes depending on the content type detected in different regions. This dynamic adaptation optimizes fidelity where needed while maintaining efficiency elsewhere, balancing the trade-off between quality and complexity.
2Loss of information
If lossless mode is applied to text and graphics regions, then information loss is reduced, but the bitstream length and processing time increase
Solution Approach 1:
Different coding qualities are applied to different regions of the video frame based on content type. Text and graphics regions receive lossless coding to preserve information, while natural image regions use lossy coding for better compression. This local differentiation reduces overall information loss in critical regions without applying computationally expensive lossless coding everywhere.
Solution Approach 2:
Instead of applying lossless coding to the entire frame, the solution applies it partially only to regions containing text and graphics. This partial action achieves the necessary information preservation for readable content while avoiding the excessive processing overhead that would result from full-frame lossless coding.
3Measurement precision
If quantization and transform steps are bypassed for text regions, then encoding precision is improved, but the computational overhead increases
Solution Approach 1:
The encoding process is segmented into different processing paths based on region type. Text/graphics regions follow a simplified path that bypasses quantization and transform steps, while natural image regions follow the full processing path. This segmentation improves precision for text without unnecessarily applying the same expensive operations to all regions.
Solution Approach 2:
The coding parameters are changed dynamically based on content type. For text regions, parameters such as quantization step size and transform application are modified or disabled to preserve precision. For natural images, standard parameters are used. This parameter adaptation achieves high precision where needed while reducing computational overhead elsewhere.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An apparatus used in video encoding comprising a processor configured to receive a video frame, select at least one region in the video frame, bypass a quantization step in encoding of the at least one region, for a current block in the at least one region generate a prediction block, subtract the current block by the prediction block to generate a residual block, and selectively bypass a transform step in encoding of the residual block.