CABAC Coding Amount Prediction for Early I_PCM Macroblock Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image coding devices face challenges in processing high-definition image data efficiently due to the uncertainty in output coded data amounts, particularly with the CABAC method, which makes it difficult to decide whether to output I_PCM data or coded data before actual binarization, leading to delayed intra-picture and motion predictions.
Innovation Solution
An image coding device that predicts the coding amount of each macroblock prior to actual CABAC processing, allowing for a decision on whether to output I_PCM data or coded data based on a threshold value, enabling earlier processing and pipelining of image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If CABAC method is used for entropy coding, then compression rate is improved, but processing speed and decision timing deteriorate due to inability to predict coding amount before actual coding
Solution Approach 1:
The patent performs preliminary binarization of DCT coefficients and motion vectors before the actual CABAC coding decision. By converting the data to binary form in advance and estimating the coding amount based on this preliminary binarization, the system can make I_PCM decisions earlier in the processing pipeline, without waiting for the actual CABAC coding to complete. This preliminary action resolves the contradiction by enabling both high compression rate (through CABAC) and earlier decision timing (through advance binarization and estimation).
2Loss of substance
If CABAC method is used for entropy coding, then compression rate is improved, but intra-picture and motion predictions are delayed
Solution Approach 1:
The patent performs preliminary binarization of DCT coefficients and motion vectors before the actual CABAC coding decision. By converting the data to binary form in advance and estimating the coding amount based on this preliminary binarization, the system can make I_PCM decisions earlier in the processing pipeline, without waiting for the actual CABAC coding to complete. This preliminary action resolves the contradiction by enabling both high compression rate (through CABAC) and earlier decision timing (through advance binarization and estimation).
Solution Approach 2:
The patent separates the binarization process from the CABAC coding process, allowing binarization to be performed independently and earlier in the pipeline. This segmentation enables the system to prepare binary data in advance while the actual CABAC coding can proceed in parallel or be deferred, thus reducing the delay to intra-picture and motion predictions while maintaining the compression benefits of CABAC.
3Reliability
If maximum coding amount restriction is enforced after CABAC, then coding compliance is ensured, but processing efficiency deteriorates due to post-coding decisions
Solution Approach 1:
The patent performs preliminary binarization of DCT coefficients and motion vectors before the actual CABAC coding decision. By converting the data to binary form in advance and estimating the coding amount based on this preliminary binarization, the system can make I_PCM decisions earlier in the processing pipeline, without waiting for the actual CABAC coding to complete. This preliminary action resolves the contradiction by enabling both high compression rate (through CABAC) and earlier decision timing (through advance binarization and estimation).
Solution Approach 2:
The patent uses feedback from the preliminary binarization stage to estimate the CABAC coding amount and make I_PCM decisions before actual coding. By feeding forward the binary data and its properties from the preliminary binarization step, the system can predict whether the coded data will exceed the maximum coding amount threshold, enabling compliance decisions to be made earlier without sacrificing reliability.
Data Source
AI summary
An image coding device in which entropy coding is applied to binarized data. An image coding unit codes image data; an entropy coding unit binarizes the coded image data and applies entropy coding to the binarized data; and a coding amount prediction unit obtains a predictive value corresponding to a coding amount of the image data on a predetermined data unit basis prior to the actual binarizing. A coding amount comparing unit compares the predictive value to a predetermined threshold value; a I_PCM decision unit outputs a selection signal indicating the image data on a predetermined data unit basis or I_PCM data, which is selected based on a result of the comparing; and an output selection unit and an output unit output the selected data indicated by the selection signal.