MVTec Software GmbH, the leading provider of standard machine vision software, presents new and outstanding machine vision features with its latest release, HALCON 18.05.
With this new version, the deep learning inference, i.e., the use of a pretrained Convolutional Neural Network (CNN), is now running on CPUs for the first time. In particular, this inference has been highly optimised for Intel®-compatible x86 CPUs. This means that a standard Intel CPU can reach the performance of a mid-range graphic processor (GPU) with a runtime of approximately two milliseconds. The operational flexibility of systems can therefore be significantly increased. For example, industrial PCs, which usually do not utilise powerful GPUs, can now be used for deep learning based classification tasks without any problems.
In addition, the new HALCON version also offers several other improvements that further increase the usability of machine vision processes. Enhanced functions for deflectometry, for instance, improve the precision and robustness of error detection for objects with partially reflective surfaces.
Developers in particular benefit from two other new features: first, they can now access HDevelop procedures not just in C++, but also in .NET via an exported wrapper – as easily and intuitively as a native function. This significantly facilitates the development process. Second, HALCON 18.05 makes it much more comfortable to work with handles. With the new version, they are automatically deleted once they are no longer required. Thus, the risk of memory leaks is significantly reduced because users no longer have to manually release unused memory.
Improved functions for bar code reading
Additionally, HALCON 18.05 also features optimised edge detection, which improves the ability to reliably read bar codes with very small line widths as well as slightly blurred codes. Moreover, the quality of the bar codes is also verified in accordance with the most recent version of the ISO/IEC 15416 standard. The new release also offers optimised functions for surface-based 3D matching: they can be used to determine the position of objects in 3D space more reliably, making development of 3D applications easier. Furthermore, a new camera model within HALCON makes it possible to correct distortions in images that were recorded with hyper-centric camera lenses. These lenses can depict several sides of an object simultaneously, thus enabling a convergent view of the test object. With this technology, users only need a single camera system for inspection and identification tasks, e.g., the inspection of cylindrical objects.
“The release of HALCON 18.05 marks another milestone for trend-setting machine vision software. In particular, the new release addresses the growing importance of AI-based technologies such as deep learning and CNNs for machine vision processes,” states Johannes Hiltner, Product Manager HALCON at MVTec.
Dr. Olaf Munkelt, Managing Director of MVTec Software GmbH, adds, “With this new version, we are pleased to offer users and developers machine vision features that are well-thought-out and extremely progressive in equal measure. The features allow our customers to further simplify their machine vision processes and raise them to a whole new level with HALCON’s handy new functions.”