Morphological and optical properties of rough surfaces

Gabriel Fricout
Abstract : Topographical, physical and chemical properties of rough surfaces are strongly involved in their visual. The understanding of the phenomena determining the appearance of a product represents a growing scientific as well as economic challenge. In this thesis, we introduce morphological and statistical features designed for characterising textures in a qualitative as well as quantitative way. These data are processed by factorial analysis methods, systematic bootstrap over-sampling and bayesian classification techniques used together to produce an automatic algorithm of surface appearance quality estimation. These theoretical advances have been applied to an industrial context. A vision system enables us to take pictures continuously during the production process of orange peel surfaces. These images taken on line are then processed to analyse the visual quality of the product.
Document type :
Theses
Complete list of metadatas

https://pastel.archives-ouvertes.fr/pastel-00001102
Contributor : Ecole Mines Paristech <>
Submitted on : Friday, April 15, 2005 - 8:00:00 AM
Last modification on : Tuesday, September 29, 2015 - 10:32:44 AM
Long-term archiving on: Friday, October 19, 2012 - 11:50:11 AM

Identifiers

  • HAL Id : pastel-00001102, version 1

Collections

Citation

Gabriel Fricout. Morphological and optical properties of rough surfaces. domain_other. École Nationale Supérieure des Mines de Paris, 2004. English. ⟨pastel-00001102⟩

Share

Metrics

Record views

438

Files downloads

147