Image Decoding Device CU Size Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The expansion and reduction of macroblock size in moving image encoding systems, as described in previous techniques, lead to a decrease in compression ratio due to increased information required to describe the size information of the macroblock.
Innovation Solution
An image encoding and decoding device that includes a CU dividing unit and a CU-size storing unit, which predicts the size of the Coding Unit (CU) using stored size information from preceding and following pictures, allowing for flexible CU size determination and reducing the information needed to describe CU size.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the macroblock size is expanded or reduced to achieve flexible prediction units, then the adaptability of prediction is improved, but the compression ratio deteriorates due to increased information required to describe the size information
Solution Approach 1:
The patent applies preliminary action by storing size information of CUs in preceding and following pictures before encoding the current picture. This pre-stored information is then used to predict CU sizes in the current picture, eliminating the need to transmit size information for each CU and thereby improving compression ratio while maintaining flexible macroblock size adaptation
2Measurement precision
If prediction is performed in a unit smaller than 16 pixels×16 pixels, then the prediction accuracy is improved, but the device complexity increases due to additional division and selection mechanisms
Solution Approach 1:
The patent applies segmentation by dividing the macroblock into smaller CUs (16×16, 8×16, 16×8, or 8×8 pixels) and performing prediction for each CU independently. This allows prediction to adapt to local image characteristics, improving prediction accuracy while the standardized division patterns keep the implementation complexity manageable
3Device complexity
If the macroblock size is limited to 16 pixels×16 pixels or smaller, then the device complexity is reduced, but the adaptability of prediction deteriorates
Solution Approach 1:
The patent resolves this contradiction by adding a temporal dimension to the problem. Instead of only considering spatial dimensions (macroblock divisions), it utilizes the time dimension by referencing and storing CU size information from preceding and following pictures. This allows the system to maintain simple 16×16 pixel macroblock processing while achieving greater adaptability through temporal prediction across multiple pictures
Data Source
AI summary
In expansion and reduction of the size of a macroblock in a moving image encoding, a compression ratio falls because an information amount for describing size information of the macroblock increases. An image decoding device includes a CU dividing unit and a CU-size storing unit. The CU-size storing unit has stored therein size information of an encoded CU in an encoding target picture to be stored and size information of all CUs in preceding and following pictures used for prediction. The CU dividing unit predicts a CU size in an encoding target position using the CU size information stored in the CU-size storing unit and determines a CU size of a final encoding target CU.