Open AccessMeasurement Strategies for Street-Level SLAM Laser Scanning of Urban Environments
Author(s) -
Aino Keitaanniemi,
Antero Kukko,
Juho-Pekka Virtanen,
Matti Vaaja
Publication year - 2020
Publication title -
the photogrammetric journal of finland/the photogrammetric journal of finland
Language(s) - English
Resource type - Journals
eISSN - 1799-2486
pISSN - 0554-1069
DOI - 10.17690/020271.1
Subject(s) - inertial measurement unit , simultaneous localization and mapping , gnss applications , point cloud , computer science , computer vision , laser scanning , artificial intelligence , mobile mapping , trajectory , remote sensing , units of measurement , satellite , mode (computer interface) , laser , global positioning system , geography , mobile robot , robot , physics , engineering , optics , human–computer interaction , telecommunications , aerospace engineering , quantum mechanics , astronomy
Data collection for street-level mapping is currently executed with terrestrial (TLS) or mobile laser scanners (MLS). However, these methods have disadvantages such as TLS requiring a lot of time and MLS being dependent on the global navigation satellite system (GNSS) and an inertial measurement unit (IMU). These are not problems if we use simultaneous localization and mapping (SLAM) based laser scanners. We studied the utility of a SLAM ZEB-REVO scanner for mapping street-level objects in an urban environment by analyzing the geometric and visual differences with a TLS reference. In addition to this, we examined the influence of traffic on the measurement strategy. The results of the study showed that SLAM-based laser scanners can be used for street-level mapping. However, the measurement strategy affects the point clouds. The strategy of walking trajectory in loops produced a 2 cm RMS and 4-6 mm mode of error even in not optimal situations of the sensor in the urban environment. However, it was possible to get an RMS under 2.2 cm and a 32 cm mode of error with other measurement strategies.