Smartphone-Based Inertial Odometry for Blind Walkers

Abstract

Pedestrian tracking systems implemented in regular smartphones may provide a convenient mechanism for wayfinding and backtracking for people who are blind. However, virtually all existing studies only considered sighted participants, whose gait pattern may be different from that of blind walkers using a long cane or a dog guide. In this contribution, we present a comparative assessment of several algorithms using inertial sensors for pedestrian tracking, as applied to data from WeAllWalk, the only published inertial sensor dataset collected indoors from blind walkers. We consider two situations of interest. In the first situation, a map of the building is not available, in which case we assume that users walk in a network of corridors intersecting at 45° or 90°. We propose a new two-stage turn detector that, combined with an LSTM-based step counter, can robustly reconstruct the path traversed. We compare this with RoNIN, a state-of-the-art algorithm based on deep learning. In the second situation, a map is available, which provides a strong prior on the possible trajectories. For these situations, we experiment with particle filtering, with an additional clustering stage based on mean shift. Our results highlight the importance of training and testing inertial odometry systems for assisted navigation with data from blind walkers.

Estimation of Spatio-Temporal Parameters of Gait and Posture of Visually Impaired People Using Wearable Sensors

“…Nonetheless, depending on leg rotations, this can also be advantageous in terms of detecting steps taken and rest periods more accurately, which can provide better insight for navigation tools. Some researchers have developed accurate methods of detecting steps using acceleration and rotation values in order to reconstruct paths for indoor navigation using deep learning models such as the uni-directional long short-term memory (LSTM) model [ 17 ]. This method for real-time assessment could be beneficial for the development of a more complex O&M rehabilitation tool.…”

Advances in Indoor Positioning and Indoor Navigation

“…Ren et al [ 17 ] compare several systems but focus on a very specific collective: people who are blind. Authors compare several algorithms using inertial sensors for pedestrian tracking, as applied to data from WeAllWalk, the only published inertial sensor dataset collected indoors from blind walkers.…”

Challenges in Acceptance of Smartphone-Based Assistive Technologies: Extending the UTAUT Model for People With Visual Impairments

“…Over the last decades, solutions have been proposed for both indoor and outdoor navigation. Specifically, indoor navigation solutions are based on inertial odometry (Ren et al, 2021), sensor-based pedestrian dead reckoning (Huang et al, 2019), indoor localization utilizing computer vision and deep learning on camera-based input or beacons readings (Koutris et al, 2022;Viset et al, 2022), as well as methods for reliably evaluating the adaptability of these solutions (Schyga et al, 2022). Likewise, outdoor navigation employs approaches incorporating both the smartphone sensors and external higher accuracy GPS receivers coupled with patent-pending novel routing algorithms (Theodorou et al, 2022a); deep learning computing vision for detecting user path obstacles, car directionality, and crosswalks near traffic lights (Chandna & Singhal, 2022;Das et al, 2021;Hsieh et al, 2021;Shelton & Ogunfunmi, 2020); and "smart" traffic lights devices for the safe passage of crossings (Theodorou et al, 2022a).…”