June 1 Report? British media said that everyone's walking posture is slightly different, and scientists are developing a technology to identify individuals including criminals by observing their gait.
The British "Daily Mail" website reported on May 29 that this system may soon be used in airports like fingerprint and eye scanning technology.
According to the report, the recognition accuracy of this non-invasive technology is about 99.3%, and it is necessary to use a pressure pad embedded in the ground of the airport. This system studies a person's gait, not the footprint shape itself.
At present, physical biometric technologies such as fingerprint, facial recognition and retinal scanning are common in the security field.
However, the report says that behavioral biometrics, including your walking posture, voice and signature, can capture the uniqueness of a person's behavior and movements.
In order to create this artificial intelligence system, researchers from the University of Manchester and the Autonomous University of Madrid in Spain collected and established the largest footstep database in history, including nearly 20,000 footstep signals from 127 people.
The research paper published in the monthly issue of Pattern Analysis and Machine Intelligence found that individuals can be accurately identified by monitoring these actions.
The leader of this research, Omar of the School of Electrical and Electronic Engineering of Manchester University? Kostyra? Reyes said: Everyone has about 24 different elements and movements when walking, which makes everyone have a unique walking pattern. ?
He said:? Therefore, monitoring these actions can be used to clearly identify or verify someone's identity, just like fingerprint identification or retinal scanning. ?
According to the report, the researchers tested their data in real security scenarios, including airport security checkpoints, workplaces and home environments. ?
In order to collect samples and data sets, the research team used ground sensors and high-definition cameras.
Reyes said: It is very challenging to pay attention to non-invasive gait recognition technology to monitor the force on the ground when walking. This is because it is extremely difficult to distinguish the subtle differences between people manually, which is why we must design a new artificial intelligence system from a new angle to overcome this challenge. ?
According to the report, other applications of this technology include intelligent measures that can identify neurodegeneration. This is another field that Reyes intends to promote his research on footprint recognition.
He said:? We are still studying how to use the first-hand footstep data provided by wide-area ground sensors in smart homes to solve the medical problems of finding markers of cognitive decline and mental illness. ?