CFA Image Noise Filtering via Motion-Selective Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise filtering techniques for digital image sequences are inadequate for portable devices due to high processing complexity and costs, leading to reduced encoding/compression efficiency and image quality.
Innovation Solution
A method for filtering digital image sequences in CFA format using a Bayer filter, which selectively applies spatial or spatio-temporal filtering based on noise estimation and motion detection, reducing noise while minimizing computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional noise filtering techniques are applied to digital image sequences, then noise reduction performance is improved, but processing complexity and computational costs increase excessively
Solution Approach 1:
The patent segments the image sequence processing into distinct phases: motion detection phase and noise filtering phase. The method selectively applies filtering only to regions identified as stationary through motion detection, rather than processing the entire sequence uniformly. This segmentation reduces overall processing complexity while maintaining noise reduction effectiveness in relevant areas.
Solution Approach 2:
The patent applies partial filtering action by using motion detection to identify only the portions of the image sequence that require noise filtering. Instead of applying filtering to all frames and regions, the method selectively filters only stationary regions, performing partial action that reduces computational burden while achieving sufficient noise reduction performance.
2Loss of information
If noise filtering is applied to improve image quality, then encoding/compression efficiency is improved, but processing time and computational resources increase
Solution Approach 1:
The patent performs motion detection as a preliminary action before applying noise filtering. By first identifying stationary regions through motion detection, the system prepares a mask that guides subsequent filtering operations. This preliminary action prevents unnecessary filtering of moving regions, reducing processing time while maintaining compression efficiency for stationary areas.
Solution Approach 2:
The method applies noise filtering partially only to stationary regions identified by motion detection, rather than processing the entire image sequence. This partial action reduces the total processing time and computational resource consumption while still achieving improved encoding/compression efficiency for the relevant stationary portions of the sequence.
3Reliability
If advanced filtering methods are used to reduce noise, then image quality is improved, but the processing cost becomes excessive for portable devices
Solution Approach 1:
The patent segments processing resources by applying different operations to different regions: motion detection is applied globally, but noise filtering is applied only to stationary regions identified by the motion detection. This segmentation reduces the total computational cost and energy consumption compared to applying advanced filtering to the entire sequence, making the solution feasible for portable devices while maintaining image quality.
Solution Approach 2:
The method uses partial filtering action by selectively applying noise reduction only to stationary regions rather than the entire image sequence. This partial application of filtering reduces processing cost and energy consumption for portable devices while still achieving sufficient image quality improvement in the regions that require it.
Data Source
AI summary
The present invention sets out to make available a method for reducing noise in an image sequence. This method can be implemented in an acquisition device such as a digital video camera or the like. The aim of this invention is attained with a method for filtering a sequence of digital images in CFA format.