User Adjustable Image Enhancement Filtering via Spatial Frequency Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image enhancement filters lack the ability for users to easily specify peak and valley filter values at desired spatial frequencies, leading to suboptimal image enhancement and increased processing time due to the need for larger kernel sizes.
Innovation Solution
A method that divides the frequency spectrum into regions with defined boundary frequencies, allowing users to specify peak and/or valley filter values, using linear or cosine functions to generate filters with specified values at desired frequencies, thereby enabling fast and simple image enhancement filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional image enhancement filters are used, then image processing can be performed, but users cannot easily specify peak and valley filter values at desired spatial frequencies
Solution Approach 1:
The frequency spectrum is divided into multiple regions with defined boundary frequencies, allowing users to specify peak and valley filter values at desired spatial frequencies within each region. This segmentation enables independent control of different frequency ranges while maintaining overall filter coherence.
Solution Approach 2:
The filter design allows dynamic adjustment of peak and valley values at specified frequencies through user input parameters. The filter adapts to user-specified constraints by dynamically calculating appropriate filter coefficients that satisfy the desired frequency response characteristics.
2Adaptability or versatility
If larger kernel sizes are used to achieve desired filter shapes, then more frequency points can be specified, but processing time increases
Solution Approach 1:
The invention changes the approach from specifying many frequency points to specifying fewer key parameters (boundary frequencies and peak/valley values). This parameter reduction allows the filter to achieve the desired frequency response with smaller kernel sizes, thereby reducing processing time while maintaining flexibility.
3Manufacturing precision
If Wiener-Helstrom filter is used, then image sharpness is improved, but the filter requires knowledge of H(u,v), N(u,v), and F(u,v) which are usually not all known
Solution Approach 1:
The invention extracts the essential frequency control requirements from the complex Wiener-Helstrom filter design. By separating the key frequency specification needs (peak and valley values at boundary frequencies) from the unknown system characteristics, the filter can achieve sharpness improvement without requiring complete knowledge of H(u,v), N(u,v), and F(u,v).
Data Source
AI summary
In an image enhancement method, computer program product, and system a set of boundary frequencies within a frequency spectrum are defined. The boundary frequencies divide the spectrum into a plurality of different spatial frequency regions. A digital image is then processed with a filter having a plurality of functions. Each function is exclusive to one of the spatial frequency regions. The functions of adjoining regions have equal values at respective boundary frequencies. A set of peak and or valley filter values of the functions at the respective boundary frequencies can be defined.