Gray level images contain a lot of surface information completing the
measurement with a conventional stylus gauge in a useful way. A signal
model for groove textures in gray level images of machined surfaces i
s proposed. The model has a parsimonious set of physically interpretab
le parameters, which can be used as a specific feature set for objecti
vely characterizing the essential geometric and statistic properties o
f groove textures. Estimation algorithms for some of the parameters ar
e given exemplarily. The performance of the model is demonstrated with
synthetic textures. Finally, two applications are discussed: the dete
ction of unwanted states in end milling processes and the automatic qu
ality assessment of honed cylinder bores.