Airport security checks could be transformed by enhancements in face recognition technology, being developed by Scottish researchers.

A team from Glasgow University have shown they can double the accuracy of computerised face recognition, by inputting an "average image" of a face, created by combining a number of photographs.

Police, customs and immigration authorities are interested in using automatic scanning systems to identify individuals at border points and in criminal investigations. But most current systems fall far short of the human brain's ability to recognise a face.

The researchers tested a system known as FaceVACS, embedded in a genealogy website MyHeritage. com, which contains more than 30,000 photos of celebrities.

When they input new photos of the celebrities only 54% were correctly identified. But, when they submitted "average faces" of the same celebrities, 100% accuracy was achieved.

Dr Rob Jenkins, lead researcher, said he was "surprised by the level of success of our trials".