Leveraging Maps and Computer Vision to Support Indoor Navigation for Blind Travelers

点击翻译按钮获取中文摘要

Leveraging Maps and Computer Vision to Support Indoor Navigation for Blind Travelers Summary The goal of this project is to develop a smartphone-based wayfinding app designed to help people with visual impairments navigate indoor environments more easily and independently. It harnesses computer vision and smartphone sensors to estimate and track the user’s location in real time relative to a map of the indoor environment, providing audio-based turn-by-turn directions to guide the user to a desired destination. An additional option is to provide audio or tactile alerts to the presence of nearby points of interest in the environment, such as exits, elevators, restrooms and meeting rooms. The app estimates the user’s location by recognizing standard informational signs present in the environment, tracking the user’s trajectory and relating it to a digital map that has been annotated with information about signs and landmarks. Compared with other indoor wayfinding approaches, our computer vision and sensor-based approach has the advantage of requiring neither physical infrastructure to be installed and maintained (such as iBeacons) nor precise prior calibration (such as the spatially referenced radiofrequency signature acquisition process required for Wi-Fi-based systems), which are costly measures that are likely to impede widespread adoption. Our proposed system has the potential to greatly expand opportunities for safe, independent navigation of indoor spaces for people with visual impairments. Towards the end of the grant period, the wayfinding software (including documentation) will be released as free and open source software (FOSS).