Noise Reducer Dynamic Amplitude Adjustment for Afterimage Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional noise reduction techniques using band-limiting filters along the time axis can cause afterimages in low-contrast images, leading to degraded picture quality due to erroneous motion detection.
Innovation Solution
A noise reducer system that includes a first subtractor, an amplitude adjuster, a second subtractor, a frame memory, and a motion detector, which adjusts the amplitude of the output signal based on a motion signal to determine a predetermined proportion of pixels as 'moving', thereby reducing afterimages and allowing the band-limiting filter to operate effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a band-limiting filter along the time axis is used to reduce noise, then noise reduction is achieved, but afterimages occur in low-contrast images due to erroneous motion detection
Solution Approach 1:
The patent applies different processing to different regions of the image based on local motion characteristics. The motion detection is performed per pixel, and the filter coefficient is adjusted locally for each pixel based on its motion state, allowing noise reduction in still regions while preventing afterimages in moving regions, even in low-contrast areas
Solution Approach 2:
The filter coefficient K is made dynamic rather than fixed. It is adjusted based on the motion detection result for each pixel, switching between a larger value (for noise reduction in still pictures) and a smaller value (to prevent afterimages in moving pictures). This dynamic adaptation resolves the contradiction by allowing optimal filtering parameters to change according to local image content
2Measurement precision
If motion detection threshold is set low to detect all motion, then more moving pictures are detected, but noise is removed even when not needed causing afterimages
Solution Approach 1:
The patent changes the filter coefficient parameter K based on motion detection results. When motion is detected (pixel determined as moving), K is set to a smaller value to prevent afterimages. When no motion is detected (pixel determined as still), K is set to a larger value for effective noise reduction. This parameter adaptation eliminates unnecessary filtering operations while maintaining accurate motion detection
3Object-affected harmful factors
If filter coefficient K is set small to prevent afterimages, then afterimages are reduced, but noise reduction effectiveness is significantly reduced
Solution Approach 1:
The filter coefficient K is made dynamic, switching between a larger value (for noise reduction) and a smaller value (for afterimage prevention) based on motion detection. This allows the system to achieve both goals: effective noise reduction in still regions and afterimage prevention in moving regions, rather than being constrained to a single fixed coefficient value
Solution Approach 2:
The image is segmented into moving and still regions at the pixel level based on motion detection. Different filter coefficients are applied to different segments: smaller K for moving pixel segments to prevent afterimages, and larger K for still pixel segments to maximize noise reduction effectiveness
Data Source
AI summary
In a noise reducer, a first subtractor 101 generates a difference N1 between an input video signal Ynow and a delayed video signal Ypre output from a frame memory 104 and obtained by delaying an output video signal Yout by a time corresponding to one screen. An amplitude adjuster 102 adjusts the amplitude of an output signal of the first subtractor 101 based on a motion signal MV. A second subtractor 103 subtracts an output N2 of the amplitude adjuster 102 from the input video signal Ynow. A motion detector 2 calculates the amount of change in the input video signal Ynow with respect to the delayed video signal Ypre for each of pixels constituting one screen, and generates, as the motion signal MV, a signal determining, as “moving,” a predetermined proportion of pixels for which the amount of change is smaller than a predetermined threshold value.