Group Leader: Dr. rer. nat. Martin Lauer
Industrial production requires inspection processes to guarantee the quality of products. These processes often include visual inspection of object geometry and surface properties. This kind of inspection is still often done by humans. However, visual inspection by humans is a highly repetitive and exhausting task so that the results of visual inspection by humans are often subjective and erratic. These limitations of human inspection and the high expenses for human labor motivate automatic visual inspection solutions.
The research group on automated visual inspection develops methods for the optical sensing of objects, surfaces, and 3-dimensional structures based on commercially available sensor and illumination devices. Our main research focus is to develop optimized approaches for the evaluation of the sensor signals. This includes methods from digital image processing, pattern recognition, and statistical signal processing, as well as model-driven approaches.
Examples of our activities are the development of a device for enhanced forensic analysis and a deflectometric approach for surface inspection and 3d-reconstruction of reflecting surfaces.