Impulsive Noise Removal Using Neighborhood Extrema
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for removing impulsive noise from digital video signals, such as median filtering, require significant computational resources due to the need for sorting neighborhood values, which hinders efficiency in compression processes.
Innovation Solution
A system and method utilizing maximum and minimum neighborhood values to replace impulsive pixels, reducing computational requirements by comparing pixel intensity to neighborhood maximum and minimum values, with optional adjustments using LOW_OFFSET and HIGH_OFFSET for scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If median filtering is used to remove impulsive noise, then noise removal effectiveness is improved, but computational complexity increases due to sorting requirements
Solution Approach 1:
The patent extracts only the necessary information (maximum and minimum values) from the neighborhood pixel set, rather than sorting all values as in traditional median filtering. This selective extraction maintains noise removal effectiveness while significantly reducing computational complexity from O(N log N) to O(N) operations.
Solution Approach 2:
The patent uses simple maximum and minimum operations instead of complex sorting algorithms. These lightweight operations provide sufficient noise removal capability without the heavy computational burden of full sorting, effectively replacing expensive operations with cheaper alternatives.
2Reliability
If median filtering is used to remove impulsive noise, then visual quality is improved, but processing speed decreases due to sorting operations
Solution Approach 1:
The patent extracts only the extreme values (maximum and minimum) from the neighborhood set, which are sufficient for detecting and removing impulsive noise. This extraction approach maintains visual quality improvement while achieving processing speeds limited only by simple comparison operations rather than full sorting.
3Measurement precision
If neighborhood values are sorted for median filtering, then accurate noise detection is achieved, but memory requirements increase
Solution Approach 1:
The patent extracts only the maximum and minimum values from the neighborhood, requiring storage of just two values per pixel instead of sorting and storing the entire neighborhood set. This dramatically reduces memory requirements while maintaining sufficient accuracy for impulsive noise detection through extreme value comparison.
Data Source
AI summary
A system and method for removing impulsive noise from a digital signal are disclosed. The system and method determines either a neighborhood maximum and/or a neighborhood minimum for a pixel of interest. The intensity of the pixel of interest is then compared to the neighborhood maximum or the neighborhood minimum to determine whether the pixel of interest should be replaced.