Abstract: This invention is directed to a method for mapping and positioning. The method has at least three steps. The first step calls for receiving sensor data based on location and a path of movement of a mobile device. The second step calls for generating, based on the sensor data, a path estimate and a location measurement vector (LMV), wherein the LMV includes one or more measurements at each of the locations from the sensor data from each of the locations. The third step entails generating, based on the sensor data, the path estimate, and the LMV, one or more unique path features.

Abstract: A system and method for mapping, tracking, positioning, and navigating in GPS-denied or GPS-inaccurate areas features accurate and automatic generation and update of a pedestrian lane map based on crowd-sourcing unique path features (UPFs) in combination with associated path estimates from a plurality of mobile devices. Searching and matching newly generated UPFs and associated path estimates, to a pedestrian lane map with known UPFs and associated lanes (or lane estimates) provides simultaneous localization and mapping (SLAM) of mobile devices in GPS-denied or GPS-inaccurate areas, including indoors, underground; dense urban streets with high buildings, natural canyons, and similar environments. UPFs are robust to noise and orientation invariant, enabling operation on low cost mobile device sensors and handling of real life human behavior. The innovative UPFs and pedestrian lane map enable route finding, generating guiding instructions, tracking, and analysis of pedestrian traffic.