Imaging technique could enhance CCTV surveillance

Story link: Imaging technique could enhance CCTV surveillance by Jan Harris

A University of Glasgow researcher demonstrated an imaging technique which could enhance the effectiveness of security and surveillance systems, at the BA Festival of Science, in York.

The technique is designed to help with facial recognition. Studies have found that both people and computers are currently poor at recognising faces and cannot reliably match a person’s face to other images of the same person.

According to Dr Rob Jenkins, who developed the technique, this is because photos of different people can be more similar than photos of the same person.

Dr Jenkins has created a method of merging multiple images of an individual to create their ‘average’ face. The technique is based on the idea that we are much better at recognising a face that is already familiar to us.

The technique takes a minimum of 12 photos of a face and combines them - producing an image that contains the ‘essence’ of that particular face and eliminates misleading information such as facial expression, posture and lighting.

Studies showed that this ‘average’ face can be much more reliably matched to other photos of the same person.

The technique has a range of possible applications - identifying terrorists from CCTV footage for example. It could also be used to create digital photos for passports or ID cards, making identity checking more reliable.