SURFACE INSPECTION AND SEGMENTATION BASED ON LOCAL REFLECTIVITY PROPERTIES
Sprache des Titels:
Visualization, Imaging And Image Processing
Image processing based surface inspection
systems operate on various clues derived froma wide range
of surface characteristics like gray–level values, color hues,
polarization features, clues from stereoscopic views to
name just a few. In this work we present image processing
algorithms that allow an image segmentation on the characteristics
the local reflection indicatrix is able to provide.
Surface areas can be segmented into specular or diffusely
reflecting zones as well as in zones determined by the direction
of the local surface normal.