Ver. 19.11 of MVTec’s HALCON machine vision software will be released on Nov. 15, 2019, according to the company.
The new version of the software will reduce the number of defect-free objects required to train the software on “positive” examples, to allow the detection of defective objects. A new, generic box finder makes it no longer necessary to train the software to recognize multiple varieties of product or shipping boxes, for example.
ECC-200 data codes in multi-core systems can be read up to three times faster and hard-to-read data codes will be read more easily after the ver. 19.11 update, according to MVTec.
HALCON ver 19.11 will have the ability to import deep-learning models in Open Neural Network Exchange (ONNX) format, which will allow use of a broader selection of deep learning networks, according to the company.
A new learning model for line scan cameras with telecentric lenses will also be added to the HALCON library.
To Learn More:
Headquarters: Munich, Germany
Product: HALCON ver. 19.11
Key Features: Generic box detection, improved data code-reading, broader selection of deep learning networks, increased line scan camera support.
What MVTec says: View more information on the HALCON software suite.