Machine vision lab has smoother approach to tile quality

The system would detect imperfections such as pinholes, crazing, rough or dull glazes, even on tiles with a texture or relief pattern, saving the industry time and money and reducing wastage.

A demonstration model is being developed at UWE's Machine Vision Laboratories, in collaboration with the University of Bath and Fima Surface Inspection Ltd. It is based on photometric stereo technology, an area where UWE is internationally recognised for its expertise.

The research is being supported by the SWRDA's Great Western Research fund, with matching funding from the academic and industry partners.

Professor Melvyn Smith, director of the Machine Vision Lab, said, “This three-year project will lead to significant advances in automating inspection of ceramic tiles. It could also have applications in other industries, where the quality of the surface is paramount, such as metals or shiny plastic components. It builds on our existing expertise in photometric stereo, and will be able to capture surface topography detail at extremely high resolution, at pixel level.”

Arwyn Roberts, chief technical officer from Fima SI Ltd said, “No commercial device exists so far for online detection and analysis of defects in material that is rapidly moving along a production line. This project with UWE and the University of Bath will allow us to become more competitive, as well as reducing the amount of waste materials.”

The new technology could also open up new markets in North America, the EU, China and India, and improve the export revenues of the South West region.