A new biometric “gait recognition” system has been developed by the National Physical Laboratory (NPL) in the UK. This means that individuals can now be identified & located by their “signature” walk.
The NPL, which cooperated with the Centre for Advanced Software Technology (CAST) and the BBC, has developed a new system through which a person’s walk could be identified. The tracking system combines a computer model of the NPL building with feeds from each on-site CCTV camera.
In each frame, the system separates an individual’s silhouette from the background. The rise & fall of head height is recorded & pattern forms are represented by a series of numbers. This is linked to the person’s identity. A computer can then produce a list of all the other sites that person has visited, & the occasions they’ve been there.
Iris scans & facial recognition systems are considered inadequate when it comes to identifying persons from a longer distance. These methods require a “cooperative subject” & high-quality imaging.
Research has been conducted by Professor Martin Hofmann from the Technical University of Munich. They have also developed a version that takes information from a person’s image, such as shadows on their clothes, which creates a much more detailed ‘signature’. Hoffman suggested that it could be used to identify bank robbers who had their faces covered.
In addition, Professor Daigo Muramatsu from Osaka University led research into how people can be identified from different camera angles. He believes the research could also have commercial applications.