AI applied to Fiber Optic connector inspection
25th June 2021
Data-Pixel has designed a new microscopy method for measuring defects and scratches using deep learning algorithms. This is a major advance in the industrial vision applied to fiber optic connectors.
Our Deep Learning software, based on AI, learns from a set of representative samples and diagnoses connector end-faces using taught criteria. This continuous learning process is carried out using images of characterized defects and scratches.
This Deep Learning solution is integrated into BLINK, our software; it establishes a scoring of defects and scratches based on a learned data set enabling a fine adjustment of the detection threshold sensitivity.