On gait recognition with smartphone accelerometer

Besides revealing useful information, like gender, age, existing impairments, the gait of every person is acknowledged to be so distinctive to allow the personal identification and it is regarded as a valid biometric authentication, similarly to fingerprinting and face recognition. Although the first analyses on the gait were conducted in laboratories with dedicated equipment, portable sensors have been exploited as they become available thanks to the technology miniaturization. Aiming at an even more unobtrusive analysis, recent proposals rely on the data acquired from the 3-axis accelerometer embedded in most of the smartphones commercially available on the market. Nevertheless the analysis must be tailored to the lower-grade accelerometer and the limited computational capability of the smartphone. This paper identifies the guidelines that the state-of-the-art research proposes for the gait recognition through a smartphone and discusses the procedures that are found as more appropriate.