MVTec builds the architecture of MVTec HALCON to remain adaptable to future developments. The modular design supports portability to new operating systems and hardware architectures and integration into new programming environments.
HALCON is available for standard PCs running Windows and Linux and for Arm-based platforms running Linux. This applies to both the HALCON library and the interactive development environment HDevelop. HALCON supports porting to additional architectures, including DSPs and specialized CPUs.
MVTec HALCON fully utilizes modern hardware. It also supports PC-based and embedded processors and exploits multi-processor and multi-core architectures, instruction set extensions such as NEON, SSE2, AVX2, and AVX512, and GPU acceleration.
For deep learning inference, MVTec provides the AI Accelerator Interface (AI²). AI² enables the use of supported AI accelerator hardware while abstracting models from specific devices. This ensures future-proof deployment and supports both MVTec-provided plugins and customer-specific accelerators. AI² accelerates deep learning applications as well as classic machine vision methods with integrated deep learning, such as HALCON Deep OCR.
Digital I/O interfaces in MVTec HALCON allow for direct integration with industrial devices. HALCON supports PLC communication via OPC UA and OPC Classic and fieldbus communication via the Hilscher-cifX interface.
HALCON operates as an OPC UA client and supports IIoT-ready production processes, including predictive maintenance.
Real-time ethernet and fieldbus communication can be implemented under Windows using Hilscher hardware.
