MVTec Updates HALCON Software

MVTec Software GmbH recently released HALCON 25.11, the company’s most recent update of the machine vision software.

The updated version adds a new deep learning feature, Continual Learning. Designed to enhance speed and efficiency in retraining deep learning models, Continual Learning requires only 5-10 images to retrain a model, which reduces time and money spent on retraining. The feature also is designed to be resistant to “catastrophic forgetting”, a situation that occurs when the neural network does not correctly recognizes the classes and features it originally learned. Finally, Continual; Learning only requires a standard CPU rather than a more expensive GPU, which enables easier integration and use of embedded and edge devices such as smart cameras and sensors.

The Continual Learning feature adds to updates included in the HALCON 25.05 version, released in April 2025, including the Deep 3D Matching tool and Deep OCR with alignment-based text correction.

Related: MVTec Releases Newest Version of HALCON Machine Vision Software

To Learn More:

Contact: MVTec Software GmbH

Headquarters: Munich, Germany

Product: HALCON 25.11 machine vision software

Key Features: New Continual Learning Tool, Deep OCR tool, Deep 3D Matching tool

What MVTec Software GmbH says: View more information on HALCON 25.11 machine vision software

Share your vision-related news by contacting Linda Wilson, Editor in Chief, Vision Systems Design.