Chinese police uses gait recognition for identification

The police in Beijing and Shanghai have begun to use a new form of surveillance. The gait recognition technology analyzes the body shapes and ways people walk to identify them, even if their faces are hidden from the camera.

The gait recognition software is part of an advance in China towards the development of artificial intelligence and data-driven surveillance.

On their website, the Chinese technology startup Watrix explains that gait functions with a low-resolution video are remotely obtainable and recognizable compared to other biometrics such as face, iris, palm print and fingerprint. With the features of the contactless, far-reaching, transparent recognition range and the difficult to disguise gait recognition, it closes the gap in the market for remote identification in the public security industry. “You don’t need people’s cooperation for us to be able to recognize their identity,” Huang Yongzhen, the CEO of Watrix, said in an interview. “Gait analysis can’t be fooled by simply limping, walking with splayed feet or hunching over, because we’re analyzing all the features of an entire body.”

Watrix’s software extracts a person’s silhouette from the video and analyzes their movements to create a model of the person’s gait. However, it is not yet able to identify people in real time. Users must upload videos to the program. Yet no special cameras are needed. The software can use footage from regular surveillance cameras to analyze the gait.

The technology is not new. Scientists in Japan, the UK and the U.S. Defense Information Systems Agency have been researching gait detection for over a decade. Professors from the University of Osaka have been working with the Japanese National Police Agency since 2013 to pilot the gait recognition software.