Image Quality Evaluation Using Gabor Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Evaluating the quality of printed images on diverse and complex surfaces, such as absorbent articles, is challenging due to varying surface materials, shapes, and seam patterns, which affects image clarity and visibility, necessitating an objective method for quality assessment.
Innovation Solution
The proposed solution involves using Gabor filters to compare original images with printed or simulated images, determining a quality value based on pixel-by-pixel differences between filtered signals, and optimizing surface attributes and image characteristics to improve image quality, with a quality score on a 0-10 scale.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If images are printed on diverse surfaces with varying materials, shapes, and seam patterns, then the versatility and adaptability of the printing system is improved, but the image clarity and visibility deteriorate due to surface variations
Solution Approach 1:
The patent creates a scanned copy of the printed image on the complex surface and compares it pixel-by-pixel with the original image data. This copying approach allows the system to evaluate image quality without physically altering the printed surface, accommodating various surface types while maintaining evaluation accuracy through digital comparison of corresponding regions.
Solution Approach 2:
The patent changes the evaluation parameters by filtering images through Gabor filters with specific center frequencies and orientations before comparison. This parameter transformation adapts the image data to account for surface variations, enabling accurate quality assessment across diverse printing surfaces by normalizing the comparison through frequency-domain processing.
2Adaptability or versatility
If manual evaluation methods are used to assess printed image quality, then the flexibility and adaptability to different surfaces is improved, but the time consumption and labor requirements increase
Solution Approach 1:
The patent replaces manual visual evaluation with an automated computer-based system that processes images through Gabor filters and calculates pixel-by-pixel differences. This substitution eliminates human labor while maintaining the flexibility to handle diverse surface types, significantly reducing evaluation time through computational automation rather than mechanical inspection.
Solution Approach 2:
The system performs self-service evaluation by automatically comparing the printed image against the original using digital signal processing. The automated algorithm independently assesses quality metrics without requiring human intervention, enabling continuous and rapid evaluation across multiple surfaces while maintaining adaptability to different printing conditions.
3Measurement precision
If the quality evaluation system processes images through multiple Gabor filters with different orientations, then the measurement precision of image quality is improved, but the computational complexity increases
Solution Approach 1:
The patent segments the image processing task by applying separate Gabor filters for different orientations (horizontal, vertical, diagonal) and then combining the results. This segmentation allows the system to analyze specific directional features independently, improving measurement precision for anisotropic surface variations while managing computational complexity through modular processing of orientation-specific information.
Data Source
AI summary
Techniques for evaluating the quality of a an image on a printing surface. The techniques generally includes receiving a first signal corresponding to the original image and a second signal corresponding to the rendition of the original image. The techniques further include filtering both signals using a common set of filters to extract at least partial contours of the original image and of its rendition and to determine a quality value of the rendition of the original image based on a comparison between the filtered images in the frequency domain.