MVTec announces the new version 19.05 of its software Halcon for May 31, 2019. The release will further enhance processes in machine vision with a range of new and updated features. The team of experts at MVTec has consistently further developed the seminal deep learning technologies. A highlight: With the new version, the deep learning inference can also be executed on CPUs of the common Arm® processor architecture. This allows customers to leverage the latest deep learning technologies on standard embedded devices. In addition, the deep-learning-based object detection has improved: The method, which clearly locates and identifies objects in the image by means of circumscribing rectangles (so-called bounding boxes), now also optimally recognizes the orientation of the objects.

Improvements in matching and code reading

Further, new features in Halcon 19.05 optimize the matching processes: Halcon's form-based matching makes it possible, for example, to specifically define model areas in which no contours should occur. This leads to more robust matching results in the context of repetitive structures.
In addition, the new HALCON release also offers some improvements in surface-based 3D matching: this allows the user to better control the quality of 3D edges via additional parameters. This leads to even more robust matches, especially in noisy 3D data.

Prerelease during the Innovation Day

During the "MVTec Innovation Day" on Tuesday, 14.05.2019, in Munich, it is possible to take a look at the new version in advance.