MVTec Software GmbH has announced the release of HALCON 21.05, the newest update for the machine vision software suite.
A major highlight of the update is the Generic Shape Matching feature, that reduces the number of operators required and also unites HALCON’s shape matching methods into a single set of operators. The changes are designed to allow users to integrate shape-matching-related features more easily.
The Deep Learning Framework is another new HALCON feature added with version 21.05. This feature enables users to create their own models within the software, breaking the dependence on pretrained networks or frameworks from third-party suppliers. HALCON also now supports the NIVIDIA TensorRT inference engine, designed to lower memory consumption.
Improved features include optimization of the integrated subpixel bar code reader for low-resolution codes, improved performance and usability for deep-learning-based OCR tools, and increased user-friendliness for the HDevelop integrated development environment.
Share your vision-related news by contacting Chris Mc Loone, Editor in Chief, Vision Systems Design