About Automated Image Analysis
In materialography, automated image analysis most often refers to digital quantitative evaluation of a microstructure image.
Quantitative measurements are typically length, width, and area, and are used for the evaluation of materialographic characteristics such as grain size, inclusions, layers, and phases or other constituents.
Digitalization of images enables the handling of high volumes of data, in order to ensure a statistical basis.
Software automation
Software automation eliminates the time-consuming process of manual identification, counting, or comparison. This can be facilitated, principally, in two different ways:
- Advanced universal software which is typically used by experts for a large variety of analyses
- Simpler and dedicated software that is optimized to a single or a few standardized analyses
How to Do Automated Image Analysis
Using automated image analysis to measure the dimensions of microstructural features includes six main process steps, which should be carried out and, if relevant, aligned with the applied standard.
1. Specimen selection
Choosing a representative specimen that includes the information of interest
2. Specimen Preparation
Ensuring an artifact free surface which will not disturb the automatic image analyses systems
3. Image illumination and filtration
Enhance contrast and ensure focus
4. Image digitization
Transforming the analog live image into a digitalized capture
5. Image processing
Remove disturbing information – improve image with software filters to improve contrast, for example
6. Image measurements – quantification, calculations, and result
Output from the complete analysis process