Evaluation of outdoor mapping accuracy
Ordenes Jara, Ivan Adolfo (2025)
2025
Automaatiotekniikan DI-ohjelma - Master's Programme in Automation Engineering
Tekniikan ja luonnontieteiden tiedekunta - Faculty of Engineering and Natural Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2025-03-06
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202503052590
https://urn.fi/URN:NBN:fi:tuni-202503052590
Tiivistelmä
Accuracy, defined as the degree to which a measurement, calculation, or specification aligns with the correct value or standard, is a crucial factor in robotics, particularly in autonomous mapping. For autonomous machines, map accuracy directly influences the precision of localization and navigation within mapped environments. Maps are generated based on data collected from various sensors onboard mobile platforms. This data, which reflects environmental features, is processed through Simultaneous Localization and Mapping (SLAM) algorithms to create accurate maps.
This thesis investigates the impact of various factors—such as sensor calibration, platform speed during data collection, and environmental feature density—on mapping accuracy. To conduct this study, a high-accuracy reference point cloud was obtained for a specific test area from a professional geographical survey company, serving as the ground truth. Maps generated from this area are compared against this reference data using point cloud analysis performed in CloudCompare, an open-source tool designed for detailed point cloud editing, manipulation, and comparison. Additionally, the study employs NDT-Map analysis as a secondary metric, focusing on the evaluation of active cells within the NDT map to provide a quantitative comparison.
The outcomes of this research will offer insights into the key parameters affecting mapping accuracy in mobile robotics and propose guidelines for improving map precision through optimized data collection processes and calibration.
This thesis investigates the impact of various factors—such as sensor calibration, platform speed during data collection, and environmental feature density—on mapping accuracy. To conduct this study, a high-accuracy reference point cloud was obtained for a specific test area from a professional geographical survey company, serving as the ground truth. Maps generated from this area are compared against this reference data using point cloud analysis performed in CloudCompare, an open-source tool designed for detailed point cloud editing, manipulation, and comparison. Additionally, the study employs NDT-Map analysis as a secondary metric, focusing on the evaluation of active cells within the NDT map to provide a quantitative comparison.
The outcomes of this research will offer insights into the key parameters affecting mapping accuracy in mobile robotics and propose guidelines for improving map precision through optimized data collection processes and calibration.