Adaptive 3D Transform Thresholding for Video Denoising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video noise reduction methods suffer from high computational complexity and visual blurring, particularly when dealing with moving objects, as they often require temporal filtering to prevent trailing artifacts.
Innovation Solution
The implementation of a video denoising system that uses motion-compensated 3D spatio-temporal transforms with thresholding in a 3D transform domain, decoupling block matching from the overlapped block processing loop, and reducing accumulation buffers from 3D to 2D+n, enabling real-time processing and reduced complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If temporal filtering is employed for greater filtering capability, then noise reduction is improved, but visual blurring increases
Solution Approach 1:
The patent segments the video processing into separate spatial and temporal filtering stages. Spatial filtering is performed first on individual frames to reduce noise, then temporal filtering is applied to the spatially filtered results. This segmentation allows each filter to optimize for its specific domain, reducing overall blurring while maintaining noise reduction effectiveness.
Solution Approach 2:
The patent applies spatial filtering as a preliminary action before temporal filtering. By pre-processing each frame to remove noise and enhance edges, the subsequent temporal filtering operates on cleaner data, reducing the likelihood of introducing blurring artifacts while still benefiting from temporal correlation for noise reduction.
2Measurement precision
If motion compensation is applied to prevent trailing artifacts, then filtering accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent implements dynamic motion compensation where motion vectors are estimated and applied adaptively during the filtering process. The motion compensation is dynamically adjusted based on the content being processed, allowing the system to maintain high filtering accuracy for moving objects while reducing computational complexity by applying compensation only where necessary rather than uniformly across the entire frame.
3Reliability
If 3D spatio-temporal transforms are used for denoising, then noise reduction is improved, but processing time increases
Solution Approach 1:
The patent segments the 3D spatio-temporal transform into separate 2D spatial transforms and 1D temporal transforms. By performing 2D transforms on individual frames first and then applying 1D temporal transforms to the results, the computational complexity is reduced compared to a full 3D transform, while still achieving effective noise reduction through the combined spatio-temporal processing.
Data Source
AI summary
In one method embodiment, receiving matched frames; forward transforming co-located blocks of the matched frames; and thresholding the transformed co-located blocks corresponding to a subset of the matched frames in at least one iteration.