Measurement using artificial intelligence
Image analysis is carried out using a unique deep learning algorithm which recognizes indentations even on low-contrast surfaces - in short, effortless repeatability and reproducibility for a wide range of operators.
- To date, more than 850,000 indentations on different surfaces and materials have been manually measured and fed into the deep learning algorithm.
- When performing an optical measurement, the algorithm mimics human behavior as if manually measured. This allows "difficult-to-detect indentations", which would not be detected by conventional image analysis, to be measured automatically.
- Specimen preparation for Vickers, Knoop and Brinell testing can be reduced to a minimum, also eliminating the need for additional manual remeasurement.