Machine Vision Software in Theory and Practice

Experts from MVTec will use live demos at the MVTec booth to show how these technologies support machine vision applications in practice. For example, a deep-learning-based bin-picking application, the Anypicker, will be presented at the booth. Here, a robotic system uses MVTec Halcon to pick up arbitrary objects with unknown shapes. The application combines 3D vision and deep learning for the first time with the goal of robustly recognizing gripping surfaces. In contrast to typical bin-picking applications, there is no need to teach object surfaces.

Another live demonstration will show how the easy-to-use image processing software Merlic uses Global Context Anomaly Detection technology to inspect electronic components on printed circuit boards. Logical anomalies are detected - both local, smaller defects such as scratches, as well as large-scale defects such as slipped labels. The technology introduced in the Merlic 5.2 release can be used for completeness checks and defect detection as part of quality control.

In addition, MVTec experts will give two presentations on the trendy topics of embedded vision and deep learning at the booth. One of them will deal with the development of an embedded vision application with Halcon using Microsoft.NET and the HDevengine. The second presentation will focus on the analysis of logical and structural anomalies with Merlic.