Defect detection on leather by oriented singularities
This paper presents a system for leather inspection based upon visual textural properties of the material surface. Defects are isolated from the complex and not homogeneous background by analyzing their strongly oriented structure. The patterns to be analyzed are represented in an appropriate parameter space using an optimization approach: in this way a parameter vector is associated to each different textured region in the original image. Finally a filter process, based upon knowledge about the parameter vectors representing the leather without defects, detects and classifies any abnormality.
Subject termsleather inspection texture analysis texture classification oriented structures parameter space vectorial bases
Unable to display preview. Download preview PDF.
- 1.A.Branca, M.Tafuri, G.Attolico, A.Distante “Directionality Detection in Compositional Textures” International Conference on Pattern Recognition, August 1996.Google Scholar
- 3.Mallat, S. and Hwang, W.L. “Singularity Detection and Processing with Wavelets”, Technical Report March 1991.Google Scholar