Automatische Erkennung von Fehlern in Holzoberflaechen.

A method for the automatic recognition of defects in wood is described. The method has been implemented in software on the Visual Interpretation System for Technical Applications (VISTA), which is currently being developed at the IITB. Tests have been performed by means of recognition experiments. Wood experts have confirmed that the regocnition results are encouraging enough in order to proceed with the investigation of a larger sample set. The method will work in real time provided it will be implemented on VISTA in hardware. Hardware modules for some of the algorithms are currently under development. (IITB)