Image Quality Measurement via Colorfulness and Sharpness Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring color image quality are either time-consuming and subjective, or they fail to correlate with human visual perception, especially under varying lighting and distortion conditions, making them unsuitable for real-time applications.
Innovation Solution
The system assesses color image quality by combining pixel-level and regional attributes such as colorfulness, sharpness, and contrast, using algorithms like the modified Naka-Rushton algorithm for color correction and edge detection, to provide a robust, distortion-independent, and lighting-robust evaluation correlated with human perception.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If subjective evaluation (MOS) is used to measure color image quality, then the measurement correlates with human perception, but the process is time-consuming and inappropriate for real-time applications
Solution Approach 1:
The patent replaces the mechanical human evaluation process with an automated computational system that calculates image quality metrics using algorithms processing colorfulness, sharpness, and contrast attributes, enabling real-time objective measurement while maintaining correlation with human perception
Solution Approach 2:
The patent transforms the subjective quality assessment into objective parameter-based measurement by quantifying colorfulness, sharpness, and contrast attributes with specific mathematical formulas, allowing automated computation that correlates with human visual perception
2Productivity
If no-reference objective image quality metrics are used, then real-time processing is enabled, but they fail to correlate with human visual perception under varying lighting and distortion conditions
Solution Approach 1:
The patent applies local quality assessment by evaluating regional attributes (colorfulness, sharpness, contrast) in different image regions separately, then combining them to capture local variations in lighting and distortion that affect human perception
Solution Approach 2:
The patent creates a composite quality metric by integrating multiple attributes (colorfulness, sharpness, contrast) with different weighting factors that reflect their relative importance to human perception, producing a comprehensive quality measure that correlates with human evaluation
3Extent of automation
If reference-based objective measurement is used, then automated processing is enabled, but it fails when no reference image is available for comparison
Solution Approach 1:
The patent extracts the reference image requirement from the measurement process by developing a no-reference methodology that directly assesses image quality attributes without comparison, enabling automated processing in scenarios where reference images are unavailable
Data Source
AI summary
An image/video may be analyzed to determine quality of its attributes, including local, global and pixel colorfulness, sharpness, and contrast to obtain an image quality measure. Invented quality may be obtained for captured images or videos and compared to a database or reference value of quality measures to identify quality of products, component anomalies, and product matches.