Digital Radiography Image Contrast Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The transition from traditional silver-halide film radiography to digital radiography in industries like aluminum casting is hindered by the lack of digital reference images, due to differences in spatial resolution and dynamic range between film and digital systems, making direct comparisons between film and digital radiographs challenging.
Innovation Solution
The development of methodologies and systems to normalize contrast and set brightness between digital images, allowing for the use of electronic digital reference images with non-film x-ray inspection systems, enabling comparisons regardless of the digital modality or bit depth, and facilitating the grading of discontinuities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If digital radiography is implemented to reduce costs and time, then productivity and cost efficiency improve, but the lack of digital reference images and differences in image characteristics prevent accurate defect grading
Solution Approach 1:
The patent creates digital reference images by capturing film reference radiographs with a digital detector, effectively copying the established film-based reference standards into the digital domain. This allows digital radiography to use the same reference images that have been validated for defect grading, ensuring measurement precision while enjoying the productivity benefits of digital systems.
Solution Approach 2:
The patent adjusts imaging parameters including pixel spacing, contrast, and brightness to match the characteristics of film reference radiographs. By changing these parameters, the digital system reproduces the visual characteristics of film images, enabling accurate defect severity grading despite the fundamental differences between film and digital detection mechanisms.
2Adaptability or versatility
If film reference radiographs are used for digital radiography, then existing reference standards can be utilized, but differences in spatial resolution and dynamic range between film and digital detectors make direct comparison inadequate
Solution Approach 1:
Instead of attempting to adapt film images to digital systems, the patent captures the film reference radiographs themselves using a digital detector. This creates true digital copies that preserve the original reference standards while being in a format suitable for digital radiography systems.
Solution Approach 2:
The patent systematically adjusts digital imaging parameters including pixel spacing, contrast, and brightness to match the characteristics of the captured film reference radiographs. This ensures that the digital reference images maintain the same quality and diagnostic value as the original film-based standards.
3Difficulty of detecting and measuring
If contrast is increased to improve discontinuity visibility, then defect detection sensitivity improves, but the apparent severity level shifts and classification accuracy deteriorates
Solution Approach 1:
The patent establishes standardized contrast and brightness parameters for digital reference images that have been calibrated to produce consistent severity level classifications. By controlling these parameters, the system maintains both defect visibility and classification accuracy, preventing the shift in apparent severity levels that occurs with arbitrary contrast adjustments.
Data Source
AI summary
Methodologies and systems to compare images with different levels of contrast are provided. Contrast is normalized between the images with different contrast levels and brightness is set. When normalizing contrast a derivative of gray level is determined for a first digital image having a first contrast level, and a derivative of gray level is determined for a second digital image having a second contrast level that is greater than the first contrast level. A ratio of the derivative of gray level for the first digital image to the derivative of gray level for the second digital image is determined, and the derivative of gray level for the first digital image is equalized with the derivative of gray level for the second digital image. Brightness of at least one image may be set automatically, such as by calculating an average pixel value excluding background and text, or manually.