Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest Cover
A digital model of the ground surface has many potential applications in forestry. Nowadays, Light Detection and Ranging (LiDAR) is one of the main sources for collecting morphological data. Point clouds obtained via laser scanning are used for modelling the ground surface by interpolation, a proces...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-04-01
|
| Series: | ISPRS International Journal of Geo-Information |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2220-9964/9/4/224 |
| _version_ | 1797571270227787776 |
|---|---|
| author | Mihnea Cățeanu Arcadie Ciubotaru |
| author_facet | Mihnea Cățeanu Arcadie Ciubotaru |
| author_sort | Mihnea Cățeanu |
| collection | DOAJ |
| description | A digital model of the ground surface has many potential applications in forestry. Nowadays, Light Detection and Ranging (LiDAR) is one of the main sources for collecting morphological data. Point clouds obtained via laser scanning are used for modelling the ground surface by interpolation, a process which is affected by various errors. Using LiDAR data to collect ground surface data for forestry applications is a challenging scenario because the presence of forest vegetation will hinder the ability of laser pulses to reach the ground. The density of ground observations will be therefore reduced and not homogenous (as it is affected by the variations in canopy density). Furthermore, forest areas are generally present in mountainous areas, in which case the interpolation of the ground surface is more challenging. In this paper, we present a comparative analysis of interpolation accuracy for nine algorithms, which are used for generating Digital Terrain Models from Airborne Laser Scanning (ALS) data, in mountainous terrain covered by dense forest vegetation. For most of the algorithms we find a similar performance in terms of general accuracy, with RMSE values between 0.11 and 0.28 m (when model resolution is set to 0.5 m). Five of the algorithms (<i>Natural Neighbour</i>, <i>Delauney Triangulation</i>, <i>Multilevel B-Spline</i>, <i>Thin-Plate Spline</i> and <i>Thin-Plate Spline by TIN</i>) have vertical errors of less than 0.20 m for over 90 percent of validation points. Meanwhile, for most algorithms, major vertical errors (of over 1 m) are associated with less than 0.05 percent of validation points. Digital Terrain Model (DTM) resolution, ground slope and point cloud density influence the quality of the ground surface model, while for canopy density we find a less significant link with the quality of the interpolated DTMs. |
| first_indexed | 2024-03-10T20:37:56Z |
| format | Article |
| id | doaj.art-8eace45f9c1340619fd6e7ed05a651ee |
| institution | Directory Open Access Journal |
| issn | 2220-9964 |
| language | English |
| last_indexed | 2024-03-10T20:37:56Z |
| publishDate | 2020-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | ISPRS International Journal of Geo-Information |
| spelling | doaj.art-8eace45f9c1340619fd6e7ed05a651ee2023-11-19T20:55:02ZengMDPI AGISPRS International Journal of Geo-Information2220-99642020-04-019422410.3390/ijgi9040224Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest CoverMihnea Cățeanu0Arcadie Ciubotaru1Department of Forest Engineering, Faculty of Silviculture and Forest Engineering, Transilvania University, 500123 Brasov, RomaniaDepartment of Forest Engineering, Faculty of Silviculture and Forest Engineering, Transilvania University, 500123 Brasov, RomaniaA digital model of the ground surface has many potential applications in forestry. Nowadays, Light Detection and Ranging (LiDAR) is one of the main sources for collecting morphological data. Point clouds obtained via laser scanning are used for modelling the ground surface by interpolation, a process which is affected by various errors. Using LiDAR data to collect ground surface data for forestry applications is a challenging scenario because the presence of forest vegetation will hinder the ability of laser pulses to reach the ground. The density of ground observations will be therefore reduced and not homogenous (as it is affected by the variations in canopy density). Furthermore, forest areas are generally present in mountainous areas, in which case the interpolation of the ground surface is more challenging. In this paper, we present a comparative analysis of interpolation accuracy for nine algorithms, which are used for generating Digital Terrain Models from Airborne Laser Scanning (ALS) data, in mountainous terrain covered by dense forest vegetation. For most of the algorithms we find a similar performance in terms of general accuracy, with RMSE values between 0.11 and 0.28 m (when model resolution is set to 0.5 m). Five of the algorithms (<i>Natural Neighbour</i>, <i>Delauney Triangulation</i>, <i>Multilevel B-Spline</i>, <i>Thin-Plate Spline</i> and <i>Thin-Plate Spline by TIN</i>) have vertical errors of less than 0.20 m for over 90 percent of validation points. Meanwhile, for most algorithms, major vertical errors (of over 1 m) are associated with less than 0.05 percent of validation points. Digital Terrain Model (DTM) resolution, ground slope and point cloud density influence the quality of the ground surface model, while for canopy density we find a less significant link with the quality of the interpolated DTMs.https://www.mdpi.com/2220-9964/9/4/224airborne laser scanningdigital terrain modelsurface interpolationLiDAR |
| spellingShingle | Mihnea Cățeanu Arcadie Ciubotaru Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest Cover ISPRS International Journal of Geo-Information airborne laser scanning digital terrain model surface interpolation LiDAR |
| title | Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest Cover |
| title_full | Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest Cover |
| title_fullStr | Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest Cover |
| title_full_unstemmed | Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest Cover |
| title_short | Accuracy of Ground Surface Interpolation from Airborne Laser Scanning (ALS) Data in Dense Forest Cover |
| title_sort | accuracy of ground surface interpolation from airborne laser scanning als data in dense forest cover |
| topic | airborne laser scanning digital terrain model surface interpolation LiDAR |
| url | https://www.mdpi.com/2220-9964/9/4/224 |
| work_keys_str_mv | AT mihneacateanu accuracyofgroundsurfaceinterpolationfromairbornelaserscanningalsdataindenseforestcover AT arcadieciubotaru accuracyofgroundsurfaceinterpolationfromairbornelaserscanningalsdataindenseforestcover |