Image Processing Apparatus Phase Change Noise Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image processing apparatuses currently adjust both meaningful and random noise-induced subtle motion changes, leading to deteriorated image quality due to the adjustment of random noise mixed in moving images.
Innovation Solution
An image processing apparatus that detects phase changes in multiple directions at different resolutions and estimates reliability based on temporal amplitude changes, multiplying the detected phase changes by reliability values to adjust only meaningful subtle motion changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If temporal filter is applied to detect subtle motion change based on phase change, then meaningful subtle motion change can be detected, but random noise adjustment is also performed causing image quality deterioration
Solution Approach 1:
The patent applies different processing strategies to different components of phase change: meaningful phase changes (correlated across directions) are enhanced while random noise (uncorrelated) is suppressed. By treating different types of phase changes with different quality factors, the system achieves local quality differentiation that resolves the contradiction between detection accuracy and image quality.
Solution Approach 2:
The patent uses feedback through cross-directional verification: phase changes in one direction are verified against phase changes in other directions. This feedback mechanism allows the system to distinguish meaningful motion (consistent across directions) from random noise (inconsistent), thereby improving both detection accuracy and maintaining image quality.
2Productivity
If phase change adjustment is performed in all directions uniformly, then all subtle motion changes are detected, but random noise from thermal effects is also adjusted
Solution Approach 1:
The patent assigns different quality factors to phase changes based on their directional correlation properties. Phase changes that show consistency across multiple directions receive higher quality factors and are adjusted more strongly, while uncorrelated phase changes (random noise) receive lower quality factors and are adjusted less. This local quality differentiation enables selective processing that improves productivity without amplifying harmful random noise.
Solution Approach 2:
The patent changes the quality factor parameter dynamically based on the correlation characteristics of phase changes across different directions. By adjusting this parameter according to the measured consistency of phase changes, the system optimizes the balance between detecting meaningful motion and suppressing random noise, thereby improving productivity while reducing harmful noise adjustment.
Data Source
AI summary
An image processing apparatus includes a change detection unit configured to detect phase changes in multiple predetermined directions from among phase changes in a luminance image in units of mutually different resolutions, and a reliability estimation unit configured to estimate reliability of the detected phase change based on temporal amplitude change information in the multiple directions determined in the luminance image. The reliability estimation unit may estimate the reliability using an amplitude change of multiple resolutions and using a value of an amplitude change equal to or greater than a predetermined threshold value among images having multiple resolutions. The reliability may become a greater value as the amplitude change becomes larger.