Image Encoding with Boundary Filtering and Transform Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding methods, such as MPEG-2, AVC/H.264, MPEG-H, and HEVC/H.265, do not effectively utilize the correlation between neighboring pixels in inter-prediction, leading to suboptimal encoding efficiency, particularly at the boundaries of prediction regions.
Innovation Solution
Implement a low-pass filter to the prediction residual signal at block boundaries during neighboring pixel non-reference prediction, followed by an orthogonal transformation process that aligns with the transformation base to the prediction region's features, using evaluation criteria to determine the filter process and transformation type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a low-pass filter process is applied to prediction signals at block boundaries during inter-prediction, then encoding efficiency is improved by reducing prediction residual signals, but device complexity increases due to additional filtering and orthogonal transformation operations
Solution Approach 1:
The low-pass filter process is applied preliminarily to prediction signals at block boundaries before the orthogonal transformation process. This preliminary filtering action reduces high-frequency components and signal errors in advance, making the subsequent orthogonal transformation more effective at reducing prediction residual signals and improving encoding efficiency.
Solution Approach 2:
The invention changes the parameters of the prediction signal processing by introducing a low-pass filter with specific filter coefficients and applying it to prediction signals at block boundaries. This parameter change enables better utilization of neighboring block correlation information, reducing prediction errors and improving encoding efficiency without excessive complexity increase.
2Productivity
If an orthogonal transformation process with closed transformation base is applied to reduce prediction residual signals, then encoding efficiency is improved, but manufacturing precision requirements increase for the transformation process
Solution Approach 1:
The low-pass filter process is applied as a preliminary action before the orthogonal transformation. By filtering the prediction signals first and reducing high-frequency components, the subsequent orthogonal transformation with closed transformation base can more effectively reduce prediction residual signals with more relaxed precision requirements.
Solution Approach 2:
The low-pass filter acts as an intermediary process between the prediction signal generation and the orthogonal transformation. It mediates by reducing signal errors and high-frequency components, thereby enabling the orthogonal transformation to achieve better encoding efficiency with more manageable precision requirements.
3Device complexity
If prediction is performed using only intra-prediction modes, then device complexity is reduced, but encoding efficiency deteriorates due to inability to utilize temporal correlation
Solution Approach 1:
The invention merges intra-prediction and inter-prediction modes into a unified prediction framework. By combining the correlation utilization capabilities of both modes and applying the low-pass filter and orthogonal transformation processes, the system achieves higher encoding efficiency while maintaining manageable device complexity through integrated processing.
Solution Approach 2:
The prediction device is designed with multi-functionality to handle both intra-prediction and inter-prediction modes universally. The low-pass filter and orthogonal transformation processes can be applied to both prediction types, enabling the system to achieve high encoding efficiency by utilizing both spatial and temporal correlation information without requiring separate dedicated processing paths.
Data Source
AI summary
An image encoding device 1 of the present disclosure includes a neighboring pixel non-reference predictor 11 configured to generate a predicted image by a predetermined neighboring pixel non-reference prediction for each pixel signal of an original image in a block unit, a filter processor 12 configured to perform a low-pass filter process on a prediction signal located at a boundary of a block of the predicted image by using a decoded neighboring signal neighboring to the predicted image under a predetermined control, a prediction residual signal generator 55 configured to generate a prediction residual signal of the block unit by using the predicted image, an orthogonal transformer 14 configured to perform an orthogonal transformation process on the prediction residual signal of the block unit under the predetermined control, and an orthogonal transformation selection controller 25 configured to control the filter processor 12 and the orthogonal transformer 14 and generate a predetermined transformation type identification signal. An image decoding device 5 of the present disclosure performs a decoding process based on a transformation type identification signal at the time of the predetermined neighboring pixel non-reference prediction.