John Wayne had his famous swagger, Naomi Campbell is paid to sashay along the catwalk in a certain way and John Cleese can make us laugh just by taking a few elongated strides for the Ministry of Silly Walks. But it is not just the stars that have recognisable walking patterns.
In fact, we all have a signature walking style that can be identified much like a fingerprint. Realising this, engineers from Southampton’s School of Electronics and Computer Science (ECS) have been working on a computer system that can analyse the gait of criminals caught on CCTV and then compare them with those of a suspect.
Professor Mark Nixon explains that the research started in the early 1990s, when biometrics was a relatively new discipline. ‘The event where I immediately thought this form of biometric would have made an impact was the murder of Jamie Bulger,’ he says. All the police had to go on was grainy CCTV pictures of the two main suspects walking away from the crime scene. They couldn’t identify them by their faces as these were obscured, so the only characteristic they could rely on for identification purposes was their gait.
‘At ECS we looked at the Bulger footage and thought: “This is something we could investigate,” and we began to look at how individual a walk can be. By using computer modelling we were able to prove the concept in a relatively short time and have been working on perfecting the technology ever since.’
The gait work started as a student project but quickly turned into a larger body of work as emerging video and computing technology began to offer new avenues of investigation. Mark and the team were real pioneers, with only a handful of other institutions pursuing gait as a viable form of biometric. In fact, there were only two active teams at the time – Southampton and The University of Calgary, Canada.
At Southampton, a working prototype is in operation in the form of an innovative tunnel that can track and record an individual’s gait. On first appearance the tunnel is a strange patchwork of multicoloured one-foot-squares. This is a novel form of chromakey – a digital image capturing technique often used in film production where the background can be ‘erased’ from the shot, leaving just the subject behind. The gait recognition version of chromakey differs from the film version because it must track individuals wearing a large array of colours, so different colours are used for each square. ‘We use the tunnel footage alongside other images of the individual walking in an everyday environment to build up a complete picture of how the individual moves.’
The tunnel uses eight cameras to record the movement of the individual as they walk, which is then recreated digitally using novel software back in the lab. Thus the individual can be viewed from any angle once they are digitally recreated. From there the team can characterise and map the unique walking patterns. The individual’s walk can then be recorded on a database and matched to CCTV footage.
The team has steadily built up a bank of images and data on volunteers to measure the differences in the way people move and to test out their theories. Ultimately the intention is to create a national databank to demonstrate to courts that individuals do walk differently and that the evidence of identification is legally safe to use. ‘The hardest thing to model is the passing of time,’ remembers Mark. ‘It is difficult to predict how a person’s walk will change as the decades pass.’
Can a walk convict?
Mark explains that subtle differences in physical attributes such as muscle strength, bone density and visual acuity all contribute to a personal walking style. ‘Gait recognition could have significant implications for police,’ he says. ‘It’s extremely difficult for someone to disguise the way they walk. We can identify them whether they are casually strolling or sprinting from a crime scene.’ Gait recognition has already been used in a number of high profile cases in Sweden. Police there successfully used it to identify a robber involved in a bank raid in which a customer was killed.
Clearly, gait recognition has a number of key applications in the fight against both terrorism and domestic crime. Criminals could, in the future, be made to walk through similar tunnels to aid identification and the device could be deployed in airports or to secure building access. ‘The airport application is one that we are keen to explore in more detail,’ stresses Mark. ’Gait recognition is non-invasive and people can be tracked at a distance without their knowledge. As part of a larger suite of biometrics, I believe gait recognition has a major part to play in the crackdown against both crime and global terrorism.’ The ECS team is working towards a prototype for the recognition technology that could be deployed at airports. Mark believes that this could take another two years to perfect, but there is already a lot of interest being shown from around the world.
