Computer Vision

1
Computer Vision
A very powerful and relatively inexpensive method
for sensing in industrial automation
applications. Pattern identification Images
provide a rich data set for object
classification, inspection, fault detection,
security, intelligent systems Position
determination position, orientation
non-contact, at a distance 2 dimensions (2D) or 3
dimensions (3D) single shot or continuous motion
tracking for control single or multiple
cameras low to high precision (down to um)
2
Hierarchal computer vision systems are the most
common. Image processing is broken down into
steps of increasing refinement / abstraction.
The most common categories are
3
Image acquisition RGB, grey-scale,
etc. Arithmetic and logical operations , -,
AND, OR 2D filtering noise removal, contrast
adjustment, etc. Image segmentation breaks image
into similar regions / objects Morphological
operations changes the shape of objects Feature
extraction calculate features (position, etc.)
of objects Pattern classification recognize
objects, detect differences, etc.