Multiplatform Mobile Laser Scanning: Usability and Performance

Yuwei Chen, Antero Kukko, Harri Kaartinen, Juha Hyyppä, Yuwei Chen 0005Sensors 2012 · DOI

TL;DR

This paper explores the versatility of mobile laser scanning (MLS) by presenting and evaluating multiplatform solutions, including vehicle-mounted, trolley-based, boat-mounted, and backpack systems. It demonstrates how adapting the platform and settings allows MLS to be effectively applied across diverse fields like urban mapping, fluvial geomorphology, precision agriculture, and permafrost monitoring, highlighting its potential for rapid, high-density 3D data acquisition.

What Problem Does This Paper Address?

Traditional stationary laser scanning (TLS) and airborne laser scanning (ALS) have limitations in terms of efficiency, cost, and flexibility for certain applications. TLS is slow for large areas, and ALS can struggle with detailed ground-level data or in environments where aircraft access is difficult. Mobile laser scanning (MLS) offers a promising alternative by combining laser scanning with vehicle-based navigation to rapidly capture dense 3D point clouds. However, the optimal platform and configuration for MLS can vary significantly depending on the application and terrain. This paper addresses the need to demonstrate and evaluate the performance of MLS across a range of platforms and challenging environments, showcasing its adaptability and pushing the boundaries of its applicability.

Methods

The paper introduces and evaluates several MLS systems, primarily focusing on the ROAMER system, which is presented in various configurations:

  • Vehicle-mounted ROAMER: This is the standard configuration for urban mapping, utilizing a vehicle for speed and stability.
  • Trolley-based ROAMER: For areas inaccessible to vehicles (e.g., pedestrian zones, farmland), the ROAMER system is mounted on a trolley for enhanced mobility and ease of handling.
  • Boat-mounted ROAMER (BOMMS): For fluvial environments, the ROAMER is installed on a boat to acquire detailed riverine topographical data for geomorphological studies.
  • Backpack MLS (Akhka): A novel approach for highly challenging terrain where even wheeled vehicles are impractical. The Akhka system is a backpack rig designed for mobility in environments like permafrost mires, featuring a downward-pointing scanner for cross-track profiles.

All systems utilize a GNSS-IMU (Global Navigation Satellite System - Inertial Measurement Unit) for direct geo-referencing of the laser scanning data. The paper details the technical specifications of the ROAMER system, including its scanner (FARO Photon 120), navigation sensors (NovAtel SPAN GPS-IMU), and data recording capabilities. Performance evaluation involves both in situ accuracy assessments using spherical targets and comparisons with reference data (e.g., TLS, RTK-GPS).

Datasets

The paper does not explicitly list datasets used for training. However, it presents evaluation results from several real-world data acquisition campaigns:

  • Urban mapping: Data collected in Tapiola, Finland, using a trolley-mounted ROAMER.
  • Fluvial geomorphology: Data acquired on a river using the boat-mounted BOMMS.
  • Precision farming: Data from a leaching field in Toholampi, Finland, using a trolley-mounted ROAMER.
  • Forestry and Snow: Data collected in Sodankylä, Finland, using a snowmobile and all-terrain vehicle for snow cover characterization.
  • Permafrost monitoring: Data from Vaisjeaggi palsa mire in northernmost Finland, using the Akhka backpack MLS.
  • Permanent MLS test field: Data collected on a dedicated test site near Helsinki, Finland, for evaluating the ROAMER system's accuracy.

The paper also utilizes spherical reference targets for in situ accuracy assessment of the Akhka system and compares data against TLS reference scans.

Key Insights and Ideas

  • Platform Adaptability is Key: The core insight is that a single MLS system (like ROAMER) can be highly versatile by adapting its platform (vehicle, trolley, boat, backpack) and sensor settings to suit diverse environmental and application requirements.
  • Backpack MLS Opens New Frontiers: The Akhka backpack system represents a significant advancement, enabling high-performance MLS in previously inaccessible terrains, thus expanding the scope of 3D surveying into areas like permafrost monitoring and detailed ecological studies.
  • High-Density Data for Detailed Analysis: MLS consistently delivers dense point clouds, enabling detailed 3D modeling, surface analysis, and change detection that are not achievable with traditional methods like aerial imagery or lower-density ALS.
  • System Calibration is Crucial: The paper emphasizes that accurate system calibration, particularly boresight calibration between the IMU and laser scanner, is critical for achieving high accuracy. Field calibration using TLS data significantly improved the ROAMER's performance.
  • Trade-offs Exist: While MLS offers speed and efficiency, factors like GNSS visibility, terrain ruggedness, and abrupt turns can impact navigation solution quality, necessitating careful planning and potentially in situ calibration.

Key Results

  • ROAMER Performance: On the permanent MLS test field, the ROAMER achieved planimetric accuracy with a standard deviation of 11 mm (CCW) and 9 mm (CW), and elevation accuracy with a standard deviation of 20 mm. RMSE for planimetric accuracy was 25 mm (CCW) and 20 mm (CW).
  • Akhka Performance: The Akhka system demonstrated strong performance, with a 2D RMSE of 18 mm in the horizontal plane and 29 mm in elevation, resulting in a 3D RMSE of 34 mm for target locations. This is comparable to, and in some aspects slightly better than, the ROAMER.
  • Accuracy Levels: Both systems can produce dense point clouds with absolute accuracy at the level of centimeters (horizontal RMSE of 23 mm for ROAMER and 17 mm for Akhka). Short-term relative precision was estimated around 12 mm.
  • Efficiency: MLS significantly speeds up data acquisition compared to traditional methods, enabling rapid mapping of large areas and detailed analysis of changes over time. For instance, boat-based MLS surveyed a 6 km river reach in just 85 minutes.

Limitations and Future Work

  • GNSS Visibility: Obstructions from buildings, trees, and terrain can degrade GNSS signal quality, impacting navigation accuracy. Future work could focus on improving trajectory accuracy in such areas.
  • System Calibration: While field calibration improved accuracy, there's a need for more thorough analysis of systematic errors, especially for multi-temporal studies where positional discrepancies can arise. Long-range calibration observations are suggested.
  • Vegetated Surfaces: The paper notes a need for more thorough analysis of using MLS point clouds for digital elevation model generation on vegetated surfaces, as low vegetation can complicate bare-ground extraction.
  • Time-Dependent Trajectory Variations: Further study is needed on correcting time-dependent variations in the trajectory, potentially using retrieved target data.
  • Backpack System Miniaturization: The Akhka system's weight (25 kg) is noted, with ongoing work to miniaturize it for single-person operation.
  • Automatic Data Correction and Field Control: Future research will concentrate on developing automatic data correction and field control schemes, as well as advanced object modeling and surface analysis methods.

Summary updated

BibTeX
@article{chen2012multiplatform,
  author = {Yuwei Chen and Antero Kukko and Harri Kaartinen and Juha Hyyppä and Yuwei Chen},
  title = {{Multiplatform Mobile Laser Scanning: Usability and Performance}},
  journal = {Sensors},
  year = {2012},
  doi = {10.3390/s120911712},
}
Figure