Video Encoder Slice Segment Filtering for Error Propagation Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image encoding methods for ADAS systems face challenges in reducing error propagation while maintaining coding efficiency and image quality, particularly when using methods like Cyclic Intra Refresh and HEVC, which can lead to increased coding overhead and decreased image quality.
Innovation Solution
An encoder and decoder system that splits images into slice segments, encodes and decodes blocks within these segments, and applies deblocking filters and sample adaptive offsets based on control information in the header areas, allowing for precise control of filter application across slice segments to minimize error propagation without compromising coding efficiency or image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional image encoding methods (Cyclic Intra Refresh, HEVC) are used to reduce coding amount, then transmission efficiency is improved, but error propagation increases and image quality deteriorates
Solution Approach 1:
The picture is divided into multiple slice segments, and each slice segment is further divided into multiple blocks. This segmentation allows independent encoding and filtering control for each block, preventing error propagation across the entire picture while maintaining transmission efficiency. The encoder processes blocks independently within slice segments, so errors in one block do not affect other blocks.
Solution Approach 2:
The patent applies deblocking filters and sample adaptive offsets selectively to specific blocks based on local characteristics. Control information in the header area of each slice segment determines whether filters are applied to individual blocks, allowing high-quality filtering where needed while skipping filtering in areas where it would cause errors or is unnecessary, thus maintaining overall image quality while reducing error propagation.
2Manufacturing precision
If deblocking filters are applied to all blocks to improve image quality, then image quality is improved, but coding overhead increases and processing complexity increases
Solution Approach 1:
Instead of applying deblocking filters to all blocks, the patent applies filters only to specific blocks where it is beneficial. The control information in the header area of each slice segment indicates which blocks should receive filtering. This partial application reduces processing complexity and coding overhead while maintaining image quality in the blocks that require filtering.
3Reliability
If slice segments are used to reduce error propagation, then error propagation is reduced, but coding overhead increases due to multiple headers
Solution Approach 1:
The patent combines multiple control information elements into the header area of each slice segment. Instead of having separate headers for each block or separate control flags for each filtering parameter, the control information in the slice segment header consolidates multiple functions: indicating which blocks should receive deblocking filters, which blocks should receive sample adaptive offsets, and other encoding parameters. This merging reduces the total number of headers and control bits required.
Data Source
AI summary
An encoder which encodes a picture includes processing circuitry and memory. Using the memory, the processing circuitry: splits the picture into a plurality of slice segments; encodes a plurality of blocks included in each of the plurality of slice segments; reconstructs the plurality of blocks encoded; adds, for each of the plurality of slice segments, control information to a header area of the slice segment, the control information being for controlling application of a filter to the slice segment; and applies, for each of the plurality of slice segments, the filter to a block which has been reconstructed in the slice segment, according to the control information of the slice segment.