MapGENIE: Grammar-enhanced indoor map construction from crowd-sourced data

Abstract

While location-based services are already well established in outdoor scenarios, they are still not available in indoor environments. The reason for this can be found in two open problems: First, there is still no off-the-shelf indoor positioning system for mobile devices and, second, indoor maps are not publicly available for most buildings. While there is an extensive body of work on the first problem, the efficient creation of indoor maps remains an open challenge. We tackle the indoor mapping challenge in our MapGENIE approach that automatically derives indoor maps from traces collected by pedestrians moving around in a building. Since the trace data is collected in the background from the pedestrians' mobile devices, MapGENIE avoids the labor-intensive task of traditional indoor map creation and increases the efficiency of indoor mapping. To enhance the map building process, MapGE-NIE leverages exterior information about the building and uses grammars to encode structural information about the building. Hence, in contrast to existing work, our approach works without any user interaction and only needs a small amount of traces to derive the indoor map of a building. To demonstrate the performance of MapGENIE, we implemented our system using Android and a foot-mounted IMU to collect traces from volunteers. We show that using our grammar approach, compared to a purely trace-based approach we can identify up to four times as many rooms in a building while at the same time achieving a consistently lower error in the size of detected rooms.

A Survey of Enabling Technologies for Network Localization, Tracking, and Navigation

“…However, we focus on solutions that have no prior knowledge of the indoor space and output primarily the physical map and optionally the measurements allow updating both user trajectory and environment map at two reference points PlaceSLAM RSS measurements provide proximity information relative to some well yes yes Average tracking error [185] recognizable places, e.g. doors 2-10 m in a 10 min walk SignalSLAM Wi-Fi/Bluetooth RSS, 4G LTE RSRP, magnetic field, GPS reference locations, yes yes Median tracking error [186] NFC tag or QR code readings at landmarks, and PDR based on IMU data 11-14 m DPSLAM Distributed particle filter to constrain the drift of a hip-mounted smartphone yes yes Localization error 3 m [187] IMU, user needs to revisit locations periodically for enabling loop closure at final location FEKFSLAM Low complexity SLAM approximation, maintains only a single hypothesis of yes yes Localization error 4 m [188] the state, requires a loop closure detection step at every measurement epoch at final location SmartSLAM Switches between DPSLAM, FEKFSLAM and other fusion algorithms to yes yes Depends on the [188] reduce complexity and save battery, while maintaining good accuracy scenario and algorithm [189] Uses IMU and a foot-mounted piezoelectric sensor to estimate the lengths and yes X Relative error 3% (length) orientations of the hallways for relative floor mapping and 4 • (orientation) CIMLoc Uses crowdsourced data from smartphone IMU sensors to derive users' yes X Average map error < 0.4 m [190] trajectories with PDR and particle filter compared to true map MapGENIE Uses foot-mounted IMU data to generate the hallways and processes them to yes X Correctly detects 88% of the [191] estimate the remaining structure (e.g., geometry of rooms and their areas) hallways and 81% of the rooms Walkie-Markie…”

“…In this case, the only use of the created maps is to assist in the positioning method. This contrasts with the MapGENIE system, where the goal is to create indoor maps that can also be used for visualization [191]. In that approach, data collected from a foot-mounted IMU is processed to detect steps using a Zero-Velocity-Update protocol.…”

A Conceptual Framework for Indoor Mapping by Using Grammars

“…The LabGrammar can benefit the indoor mapping process by improving the accuracy of generated maps and by dramatically reducing the volume of the sensor data required by traditional reconstruction approaches, such as LiDAR point clouds (Philipp et al, 2014) and users' traces (Alzantot et al, 2012;Zhang et al, 2014). To represent the layout rules of a laboratory, we redefine its spatial structure by using a hierarchical semantic division.…”

“…Moreover, generated maps mainly contain geometric information, and very few semantic information can be detected with Lidar point clouds. The work in (Philipp et al, 2014) uses split grammars to describe the generation process of rooms. The grammar of one floor can be learned automatically and then be used to derive the layout of the other floors.…”

Modelling of Indoor Environments Using Lindenmayer Systems

“…More recent developments aim at the integration of sensor data and can be found in the adaptation of a Palladian grammar by (Khoshelham and Díaz-Vilariño, 2014) and our previous work combining a split grammar , Philipp et al, 2014 and an L-system . While (Khoshelham and Díaz-Vilariño, 2014) -building on the Palladio grammar described in (Stiny and Mitchell, 1978) -employ a small set of general rules to support reconstruction from complete lidar point clouds, their approach does not enable for a description of repeated structures and thus does not need to learn grammar parameters from data beforehand.…”