Chinese authorities use gait analysis to identify citizens on CCTV

A Chinese start-up has developed technology that can apparently identify people based on their body shape and how they walk. Cities in China are implementing new ‘gait recognition’ software that claims to be able to identify people using their silhouettes and how they walk, even if their faces are not visible.

A Chinese surveillance company, Watrix, has developed the new system for gait recognition that can identify people as far as 50 metres away, according to CEO Huang Yongzhen. It can analyse how people carry themselves and, while it is not capable of real-time recognition yet, the company claims it can search an hour’s worth of footage in 10 minutes with an accuracy rate of 94pc. There are 170m CCTV cameras in the country and that figure is tipped to grow, with more than 400m set to be installed by 2020.

The technology is already being used by police in Beijing and Shanghai. The police are apparently using the software to locate people in crowds and apprehend jaywalkers, with the end goal of developing an integrated national system of CCTV camera data. Watrix claims the technology will be helpful in tracking down criminals across the country.

Huang said: “You don’t need people’s cooperation for us to be able to recognise their identity. Gait analysis can’t be fooled by simply limping, walking with splayed feet or hunching over, because we’re analysing all the features of an entire body.”

While Watrix has made certain claims about the technology, these have not been verified by independent experts and the effectiveness of the software has not been properly quantified. The company announced last month that it had raised 100m yuan ($14.5m) to accelerate the development and sale of the tech.