Adaptive Motion Compensated Temporal Filtering for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion compensated temporal filtering (MCTF) techniques for video coding are inadequate due to noise and motion vector errors, which lead to low encoder efficiency and poor image quality.
Innovation Solution
The proposed method employs an adaptive MCTF system that selects reference frames based on video content analysis and computes weights using robust multiple measurements of image data distortions, thereby improving noise reduction and encoder efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional MCTF techniques are used for video coding, then the encoding process can be performed, but noise and motion vector errors lead to low encoder efficiency and poor image quality
Solution Approach 1:
The patent performs preliminary denoising operations before the main encoding process by selecting reference frames and computing weights in advance. This preliminary action removes noise from the video data before encoding, improving image quality and encoder efficiency by preventing noise from degrading the encoding process
Solution Approach 2:
The patent implements feedback mechanisms by comparing original frames with motion compensated reference frames, computing distortion measurements, and using these measurements to iteratively refine weight computations. This feedback loop enables the system to adaptively adjust filtering strength based on actual noise and motion characteristics
2Loss of information
If motion compensated blocks from reference frames are weighted to form filtered image data, then temporal filtering is achieved, but noise and motion vector errors remain inadequate
Solution Approach 1:
The patent changes multiple parameters to improve noise reduction and measurement precision: it adjusts reference frame selection criteria based on content type, modifies weight computation formulas to incorporate distortion variance and dispersion distribution, and adapts filtering strength based on measured noise levels. These parameter changes enable more accurate distinction between noise and intentional image content
Solution Approach 2:
The patent introduces dynamic adaptation by computing distortion measurements for each block and adjusting weights based on local characteristics rather than applying uniform filtering. The system dynamically selects reference frames and computes weights based on real-time analysis of motion vectors, noise levels, and content complexity
3Productivity
If adaptive reference frame selection and robust weight computation are implemented, then coding efficiency is enhanced, but system complexity increases
Solution Approach 1:
The patent segments the video processing into distinct functional modules: reference frame selection unit, distortion computation unit, weight computation unit, and filtering application unit. Each module handles a specific aspect of the adaptive MCTF process, making the complex system more manageable and implementable while maintaining high coding efficiency
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods, articles, and systems of image processing comprise obtaining image data of frames of a video sequence. The method also includes determining multiple reference frames of a current frame in the video sequence. The multiple reference frames each have at least one motion compensated (MC) block of image data. Also, the method then includes generating a weight that factors noise, distortion variance, and dispersion distribution between the same MC block position and the current block. Thereafter, the method includes generating denoised filtered image data comprising applying one of the weights to the image data of the motion compensated (MC) block.