Manufacturing Technology

GEFRA believes it is leading the newest development for visual inspection with the introduction of ‘deep learning’ technology into its OPTISORT testing system, which offers users completely new perspectives – especially with complex and varying imaging conditions. 

With ‘deep learning’ GEFRA reports that defective parts and good parts can be learned and classified accordingly – with these datasets trained using multi-layer neural network architecture. Based on these templates, the test items can be classified as good or defective parts. 

“‘Deep learning’ technology is making machine vision technology for automated visual inspection more accessible and capable,” reports GEFRA GmbH. “It requires a lot of computing power but high performance GPUs have a parallel architecture that is efficient for ‘deep learning’. The easy operation of this application is also ensured by our user-friendly AutoControl testing software.”

The AutoControl interactive environment provides the platform for training these models for use in machine vision applications. The AutoControl testing software delivers all the functionality needed for this task, so an operator can create and label the training image dataset; augment the image dataset, if necessary; as well as train, analyse and test the neural network model. 

“Artificial Intelligence, specifically machine learning by way of ‘deep learning’ technology, mimics how the human brain processes visual input but performs this task with the speed and robustness of a computerised system,” highlights GEFRA. “The technology works to ensure quality in manufacturing industries, controlling production costs and enhancing customer satisfaction.  The technology still benefits from conventional image processing and analysis to locate regions of interest within images, but speeds up the overall process and make it even more robust.”