Beijing firm unveils gait recognition product to help identify suspects

By Liu Caiyu Source:Global Times Published: 2018/10/26 14:09:40

This product targets suspects by using gait recognition to monitor human postures and is on display in Beijing. Photo: Liu Caiyu/GTA

Beijing-based high-tech company released a security product on Friday using gait recognition technology to search for targets, which is likely to assist China’s public security department in identifying suspects.

Developed by WATRIX, this product, called “Shuidi Shenjian” can target suspects by monitoring their posture as the suspects walk from a distance of up to 50 meters.

Different from facial recognition, the gait recognition technology is more flexible. It is capable of identifying targets from any angle regardless of if they cover their faces or wear different clothes or walk outside at night, without suspects being aware, Huang Yongzhen, CEO of the WATRIX, told the Global Times on Friday.

Huang said that the company had inked deals with public security departments in Beijing’s Fengtai district and Shanghai. The public security department in Xinjiang also expressed interest in the security product.

The portable machine, weighing about 15 kilograms, can replace manual workers by quickly spotting suspects while they are walking. Users need to upload videos onto the machine first and offer the machine an example video of the target, then the machine can complete a search through one hour of video in 10 minutes. Its accuracy is about 94 percent.

But Huang said the technology still cannot detect targets in real-time videos, which is the next step of their research.

At present, the product has been piloted in the public security system for more than 1,000 hours, being used in the detection of more than 20 cases. It had retrieved 2,000G of public security videos so far.

The gait recognition technology can also be applied in the medical field by allowing medical personnel to examine whether a patient has fallen down and it can also be applied in oil fields to seek out intruders, as well as being used in designing smart home furniture, Huang noted.