The objective of the project is to provide cost-effective colour and texture-based automatic inspection and sorting solutions for industry. Three application areas will be considered: hot steel strip, timber and food.
Novel solutions for low-delay image analysis, allowing a few tens of milliseconds from imaging to decision, will be developed and used in real-time on-line inspection demonstrators for each application area. The platform will cope with the high volumes of data associated with colour and texture inspection. The basic technology underlying high-performance personal computers will be used. Because of the vibrations of hot steel strip and the nature of the motion of food particles to be sorted, high-speed prism-based colour line-scan cameras will be developed to capture the RGB values of each pixel simultaneously. Due to the required high line-scan rates, stable illuminators with feature-enhancing radiation patterns will be designed, and an on-line colour camera calibration technique will be developed to make the colour measurements independent of longer term changes in illumination.
The consortium intends to exploit the results of the project by marketing and selling inspection systems to industrial end-users, system platforms and dedicated hardware modules to system integrators and by using the developed inspection systems in their own production.
Funding SchemeCSC - Cost-sharing contracts