Automatic detection of people with reduced mobility: promoting accessibility in the São Paulo metro

Abstract

The São Paulo Metro serves more than 40,000 users every month, and a portion of these users have limited mobility and use wheelchairs or other auxiliary devices, such as crutches and walkers. To meet this demand, subway stations have elevators that allow the user to move between the floors of the station. However, in order for the elevator to be released to the user, station operators need to visually verify that the user requesting access fits the public served. This check is currently performed in two forms: face-to-face, in which the operator goes to the elevator, or remotely, through of videos recorded by cameras installed at the entrances of the elevators. In this way, the The present project aims to implement and evaluate artificial intelligence methods for detecting automatic display of people on crutches or other assistive devices from the images recorded by video cameras at São Paulo Metro stations.(AU)