Image Processing Apparatus Adaptive Filtering Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies face challenges in efficiently eliminating noise with small signal differences between pixels, leading to increased hardware costs and reduced noise reduction effectiveness for color noise.
Innovation Solution
An image processing apparatus that extracts similar pixels by comparing signal values and performs filtering processing, with a processing-repetition number decision unit determining the number of repetitions based on predetermined conditions to effectively reduce noise while preserving image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the number of reference pixels is increased to enhance noise elimination effect, then noise reduction effectiveness is improved, but frame memory capacity and hardware size increase
Solution Approach 1:
The patent divides the image into multiple blocks and processes each block independently with a limited number of reference pixels. This segmentation allows the system to use fewer reference pixels per block while still covering the entire image, thereby reducing frame memory capacity requirements while maintaining noise elimination effectiveness.
Solution Approach 2:
The patent applies different processing approaches to different regions by dividing the image into blocks. Each block is processed with local reference pixels, allowing the system to adapt to local noise characteristics without requiring a large global reference pixel set, thus reducing hardware memory requirements while maintaining effective noise reduction.
2Reliability
If a low-pass filter is used to reduce noise, then noise reduction is achieved, but image edges and important elements are lost
Solution Approach 1:
The patent processes each block independently and allows different blocks to have different numbers of repetitions based on their local characteristics. This local quality approach enables effective noise reduction in uniform areas while preserving edges and important features in other areas, preventing information loss.
Solution Approach 2:
The patent dynamically adjusts the number of filtering repetitions based on local image characteristics such as variance and gradient magnitude. This dynamic adjustment allows the system to apply stronger filtering where noise predominates while preserving edges and important features, thereby reducing noise without losing critical image information.
3Reliability
If the number of repetitions of filtering processing is increased to eliminate wide-range noise, then noise elimination effectiveness is improved, but processing time and computational cost increase
Solution Approach 1:
The patent divides the image into multiple blocks and processes each block independently with a limited number of repetitions. This segmentation allows the system to achieve effective noise reduction across the entire image by processing smaller regions in parallel, thereby reducing total processing time while maintaining noise elimination effectiveness.
Solution Approach 2:
The patent dynamically determines the number of filtering repetitions for each block based on local characteristics such as variance and gradient magnitude. This dynamic approach allows the system to apply more repetitions only where necessary, reducing overall processing time while maintaining effective noise elimination where needed.
Data Source
Figure 1~3
Figure 4
Figure 5
AI summary
An image processing apparatus includes a filtering unit configured to extract a pixel having a high similarity to a pixel of interest by comparing signal values of the pixel of interest and its neighboring pixels in a process-target area containing the pixel of interest in a process-target image, and perform operations for filtering processing by using the extracted pixel, and a processing-repetition number decision unit configured to determine the number of repetitions of the filtering processing based on a predetermined condition held by a process-target pixel or the process-target area, the filtering processing being repeated while setting a result of the filtering processing obtained by the filtering unit as a new signal value of the pixel of interest.