Industrial vision, and in particular image classification, is a key component of many manufacturing tasks automation such as products sorting, control, identification... The vital elements of such a vision system are on one side the hardware (CCD, ligthing,...) and on the other side the software that processes the acquired image to draw control actions and alarms.

The key problem with current image classification systems is the lack of generic procedure and algorithms. Indeed, for each classification problems you need to extract manually the relevant features that will be useful to classify the image. Moreover in some cases the images processing is so complex that it is not suitable for realtime applications.

In order to bring an answer to these key issues, successful researches were achieved in partnership with the University of Liège in the area of automatic learning and computer vision. The spectacular results of this hard work have been included in PiXiT, a toolbox developped and maintain by PEPITe for the benefits of its customers.

Some highlights of PiXiT 1.0

	Build Easily and quickly images classifiers
	Use very generic methods to build the classifiers
	Manage very larges images in many formats
	Deploy very fast classifiers (suitable for demanding real time applications)
	Construct robust and relialable classifiers

Note also that this technology is remarkably adapatable to any indexed information classification (temporal signal classification, 3D images classification,...).