Machine Vision

Computer vision techniques for industrial applications

These techniques are concerned with processing visual information generated in 2D (images), 3D (images + depth) and 4D (image-depth +time: spatiotemporal visual data with depth).

The machine vision activities in the group are not only limited to using conventional visual cameras, but also using non-visual cameras (UV, near infrared and thermal), depth sensors (e.g. stereo camera and structured light imaging) hyperspectral/multispectral cameras, optical sensors (e.g. LIDAR) and radar imaging.

The major tasks of the group are recognizing robust techniques for semantic segmentation, recognition and perception.