Variable Length Encoder Truncated Unary Code Offset
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video encoding devices face a decrease in coding efficiency due to the increased code amount required for encoding offsets during the pixel adaptive offset process, especially as the screen is partitioned into finer blocks, leading to a rise in the amount of high-frequency components lost and resulting in a blurred video image.
Innovation Solution
An image encoding device that determines a classification method for the largest coding blocks, performs class classification on each pixel within these blocks, calculates offset values, and uses a variable-length encoder to encode the classification method and offset parameters using a truncated unary code, reducing the code amount needed for encoding offsets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the screen is partitioned into finer blocks for pixel adaptive offset process, then the offset accuracy is improved, but the code amount required for encoding offsets increases
Solution Approach 1:
The patent changes the encoding parameter from conventional unary code to truncated unary code for offset values. This parameter change in the encoding method reduces the average code length while maintaining the ability to represent the same range of offset values, thereby reducing the code amount without sacrificing offset accuracy.
Solution Approach 2:
The patent applies different encoding strategies to different regions based on their characteristics. By using truncated unary code specifically for offset values in the pixel adaptive offset process, the patent optimizes the code length locally for this specific parameter while leaving other encoding parameters unchanged, thus reducing overall code amount without affecting other aspects of the encoding.
2Productivity
If the compression ratio is increased, then the compression efficiency is improved, but the quality of prediction reference image degrades
Solution Approach 1:
The patent applies pixel adaptive offset process before the prediction process. By pre-compensating for block distortion in the decoded image through offset addition, the prediction reference image quality is improved in advance, which subsequently improves motion-compensated prediction efficiency and overall compression efficiency at higher compression ratios.
Solution Approach 2:
The patent uses a feedback mechanism where the decoded image is fed back through the pixel adaptive offset process and loop filter, and the corrected image is then used as reference for subsequent prediction. This closed-loop approach continuously improves reference image quality, maintaining prediction accuracy even at high compression ratios.
3Productivity
If high-frequency components are removed during compression, then the compression efficiency is improved, but the video image becomes blurred
Solution Approach 1:
The patent applies pixel adaptive offset process before compression to pre-compensate for block distortion. By addressing block effects early in the processing chain, the patent reduces the need for aggressive compression that would remove high-frequency components, thereby maintaining image sharpness while still achieving good compression efficiency.
Data Source
AI summary
A variable length encoding unit 13 variable-length-encodes an index indicating a classification method of carrying out a class classification on each coding block having a largest size, the classification method being determined by a loop filter unit 11, and also variable-length-encodes a parameter about an offset value for each class determined for each coding block having the largest size on the basis of a binarization process using a truncated unary code.