Image Processing Device Motion Vector Accuracy Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing image processing technologies, such as those using MPEG2 and AVC, face decreased encoding efficiency due to rounding processes that reduce arithmetic operation accuracy, particularly in motion compensation and weighted prediction, especially for large image frames like UHD, where macroblock sizes are not optimal and tools for handling brightness changes in fade scenes are lacking.
Innovation Solution
An image processing device and method that performs motion compensation of sub-pixel accuracy for L0 and L1, applies weighted addition, and then applies a rounding process to the arithmetic operation results, with an adjustment unit to determine and adjust the accuracy of motion vectors based on predetermined coefficients, and performs Explicit or Implicit Weighted Prediction and Intensity Compensation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If rounding process is applied in motion compensation and weighted prediction, then device complexity is reduced and ease of operation is improved, but arithmetic operation accuracy deteriorates and encoding efficiency decreases
Solution Approach 1:
The patent applies preliminary action by performing motion compensation with high precision before the rounding process, and by pre-calculating weighted prediction values with maintained accuracy. The system performs sub-pixel accuracy motion compensation and weighted prediction operations before final rounding, ensuring that critical arithmetic operations maintain high precision throughout the processing chain, thereby resolving the contradiction between ease of operation and arithmetic operation accuracy.
2Device complexity
If rounding process is applied early in motion compensation, then device complexity is reduced, but image quality degradation increases and encoding efficiency decreases
Solution Approach 1:
The patent delays the rounding process until after motion compensation and weighted prediction are completed with high precision. By performing motion compensation at sub-pixel accuracy levels and maintaining precision through the weighted prediction stage, the system ensures that rounding only occurs at the final stage, thereby minimizing image quality degradation while keeping device complexity manageable.
Solution Approach 2:
The patent implements beforehand cushioning by maintaining higher precision arithmetic operations throughout the motion compensation and weighted prediction processes. The system uses extended precision calculations and only applies rounding at the final output stage, effectively cushioning against accuracy loss that would otherwise propagate through subsequent processing stages and degrade image quality.
3Adaptability or versatility
If macroblock size is kept at 16×16 pixels for compatibility, then adaptability to existing standards is improved, but encoding efficiency decreases for large image frames like UHD
Solution Approach 1:
The patent applies dynamics by enabling flexible macroblock size adaptation while maintaining compatibility with existing 16×16 pixel standards. The system can dynamically adjust macroblock dimensions based on the specific requirements of the image content and resolution, allowing larger macroblocks for UHD content to improve encoding efficiency while retaining the ability to use traditional sizes for compatibility scenarios.
Solution Approach 2:
The patent implements parameter changes by allowing the macroblock size parameter to be adjusted based on the imaging characteristics and resolution requirements. For UHD and large image frames, the system can increase macroblock dimensions beyond the standard 16×16 pixels, thereby improving encoding efficiency and productivity while maintaining adaptability to different application scenarios through parameter flexibility.
4Productivity
If motion vector accuracy is increased for better prediction, then encoding efficiency is improved, but calculation amount increases and device complexity increases
Solution Approach 1:
The patent applies local quality by implementing differentiated motion vector accuracy strategies for different regions and scenarios within the image. Rather than uniformly increasing motion vector precision across the entire image, the system adjusts accuracy levels locally based on motion complexity, scene characteristics, and region importance, thereby improving encoding efficiency in critical areas while avoiding unnecessary calculation increases in static or simple regions.
Data Source
AI summary
The present technology relates to an image processing device and a method capable of improving encoding efficiency. In a weighted prediction process performed in an image encoding process or in an image decoding process, a motion compensation unit that performs motion compensation of sub-pixel accuracy for each of L0 and L1, a weighted addition unit that applies weighted addition to arithmetic operation results by the motion compensation unit, and a rounding processing unit that suppresses a decrease in arithmetic operation accuracy by performing a necessary rounding process once with respect to an arithmetic operation result by the weighted addition unit are provided. The present disclosure can be applied to an image processing device.