Pattern Recognition

"To understand is to perceive patterns" - Isaiah Berlin

Pattern recognition is used to analyse recurring patterns in data with the aim to detect, understand or classify the pattern.

Pattern recognition is the study of how machines can observe the environment, learn to distinguish patterns of interest from their background and other objects and make robust and reasonable decisions.  SINTEF’s goal is to find the best possible way of utilizing available sensors, processors, image analysis and pattern recognition to make decisions automatically. We focus on developing robust systems which are capable of handling changing conditions.

Pattern recognition includes disciplines like discriminant analysis, feature extraction, error estimation, cluster analysis (together sometimes called statistical pattern recognition), grammatical inference and parsing (sometimes called syntactical pattern recognition).

SINTEF's competence covers algorithm development as well as real-time software and hardware solutions used when implementing the algorithms. 

Application examples:

• Bar code reading. You see them everywhere, attached to food articles in the grocery store, on parcels you receive, on books in your local bookstore or on the CD's in your stereo shelf; the bar codes are all around us. The ordinary way to read bar codes is by using dedicated and expensive laser based bar code readers. SINTEF has comprehensive experience in using normal and often very cheap cameras instead.
  
• Face recognition
• Fingerprint recognition
• Medical applications
• Process and product inspection
• Character recognition
• Seismic analysis

Analysis of extruded aluminium profiles for Hydro Aluminium. Pattern recognition techniques were used to find similarities between cross-sections of earlier produced aluminium profiles and new ones. Based on the similarities, process parameters for the new aluminium profile could be estimated.