An inertial and QR code landmarks-based navigation system for impaired wheelchair users

Personal mobility is a key factor in independent living for elderly people and people with motor disabilities, thus indoor navigation systems are of utmost concern in Ambient Assisted Living (AAL) applications. Driving an electric powered wheelchair in domestic environments becomes difficult for people with arms or hands impairments. Moreover, people affected by tetraplegia are completely unable to operate a joystick, and must rely on input interfaces, such as eye tracking and “sip and puff”, which require tedious and repetitive tasks to be operated. Smart powered wheelchairs with autonomous navigation intelligence and their integration within AAL homes, may enhance independence and improve both the security and the perceived quality of life. Self-navigating systems combine different measurements provided by both absolute and relative sensors to improve localization accuracy. In this work, a low-cost localization system for autonomous wheelchairs, which takes advantage of Quick Response (QR) code landmarks information, is proposed. QR code is a low-cost pattern with fast readability and large storage capacity with respect to other landmarks solutions. The proposed wheelchair is equipped with an Inertial Measurement Unit (IMU) and a video camera: the inertial information, provided by the IMU, is fused with that provided by QR code recognition, thus reducing the error propagation caused by a Dead Reckoning (DR) approach. Autonomy and intelligence of the wheelchair is drastically increased by integrating within its navigation system both the knowledge about self localization and the environment (e.g. room identification). QR code landmarks are a suitable solution to store this information. This approach has been implemented and experimentally tested in an indoor scenario, demonstrating its feasibility and its good and reliable long-term performances.