Mapping subaerial sand-gravel-cobble fluvial sediment facies using airborne lidar and machine learning

Abstract. Automatic semantic segmentation of point clouds observed in a 3D complex urban scene is a challenging issue. Semantic segmentation of urban scenes based on machine learning algorithm requires appropriate features to distinguish objects from mobile terrestrial and airborne LiDAR point clouds in point level. In this paper, we propose a pointwise semantic segmentation method based on our proposed features derived from Difference of Normal and the features “directional height above” that compare height difference between a given point and neighbors in eight directions in addition to the features based on normal estimation. Random forest classifier is chosen to classify points in mobile terrestrial and airborne LiDAR point clouds. The results obtained from our experiments show that the proposed features are effective for semantic segmentation of mobile terrestrial and airborne LiDAR point clouds, especially for vegetation, building and ground classes in an airborne LiDAR point clouds in urban areas.

Download Full-text

Forest biomass estimation from airborne LiDAR data using machine learning approaches

Remote Sensing of Environment ◽

10.1016/j.rse.2012.07.006 ◽

2012 ◽

Vol 125 ◽

pp. 80-91 ◽

Cited By ~ 144

Author(s):

Colin J. Gleason ◽

Jungho Im

Keyword(s):

Machine Learning ◽

Forest Biomass ◽

Airborne Lidar ◽

Biomass Estimation ◽

Lidar Data ◽

Learning Approaches ◽

Airborne Lidar Data

Download Full-text

Classification of Photogrammetric and Airborne LiDAR Point Clouds Using Machine Learning Algorithms

Drones ◽

10.3390/drones5040104 ◽

2021 ◽

Vol 5 (4) ◽

pp. 104

Author(s):

Zaide Duran ◽

Kubra Ozcan ◽

Muhammed Enes Atik

Keyword(s):

Machine Learning ◽

Point Cloud ◽

Point Clouds ◽

Machine Learning Algorithms ◽

Airborne Lidar ◽

Aerial Photogrammetry ◽

Feature Spaces ◽

Wide Range ◽

Extract Information

With the development of photogrammetry technologies, point clouds have found a wide range of use in academic and commercial areas. This situation has made it essential to extract information from point clouds. In particular, artificial intelligence applications have been used to extract information from point clouds to complex structures. Point cloud classification is also one of the leading areas where these applications are used. In this study, the classification of point clouds obtained by aerial photogrammetry and Light Detection and Ranging (LiDAR) technology belonging to the same region is performed by using machine learning. For this purpose, nine popular machine learning methods have been used. Geometric features obtained from point clouds were used for the feature spaces created for classification. Color information is also added to these in the photogrammetric point cloud. According to the LiDAR point cloud results, the highest overall accuracies were obtained as 0.96 with the Multilayer Perceptron (MLP) method. The lowest overall accuracies were obtained as 0.50 with the AdaBoost method. The method with the highest overall accuracy was achieved with the MLP (0.90) method. The lowest overall accuracy method is the GNB method with 0.25 overall accuracy.

Download Full-text

Assessing the regional-scale distribution of height growth of Cryptomeria japonica stands using airborne LiDAR, forest GIS database and machine learning

Forest Ecology and Management ◽

10.1016/j.foreco.2021.119953 ◽

2022 ◽

Vol 506 ◽

pp. 119953

Author(s):

Katsuhiro Nakao ◽

Daisuke Kabeya ◽

Yoshio Awaya ◽

Shin Yamasaki ◽

Ikutaro Tsuyama ◽

...

Keyword(s):

Machine Learning ◽

Cryptomeria Japonica ◽

Regional Scale ◽

Height Growth ◽

Airborne Lidar ◽

Gis Database

Download Full-text

Identifying vegetation-geomorphology relationships in permafrost with airborne LiDAR, electrical resistivity tomography, seasonal thaw depth measurements, and machine learning

10.5194/egusphere-egu2020-1900 ◽

2020 ◽

Author(s):

Thomas Douglas ◽

Christopher Hiemstra ◽

John Anderson ◽

Caiyun Zhang

Keyword(s):

Machine Learning ◽

Snow Depth ◽

Electrical Resistivity Tomography ◽

Airborne Lidar ◽

Spruce Forest ◽

Snow Pack ◽

Resistivity Tomography ◽

Interior Alaska ◽

Thaw Depth ◽

Ice Content

Mean annual temperatures in interior Alaska, currently -1&#176;C, are projected to increase as much as 5&#176;C by 2100. An increase in mean annual temperatures is expected to degrade permafrost and alter hydrogeology, soils, vegetation, and microbial communities. Ice and carbon rich &#8220;yedoma type&#8221; permafrost in the area is ecosystem protected against thaw by a cover of thick organic soils and mosses. As such, interactions between vegetation, permafrost ice content, the snow pack, and the soil thermal regime are critical in maintaining permafrost. We studied how and where vegetation and soil surface characteristics can be used to identify subsurface permafrost composition. Of particular interest were potential relationships between permafrost ice content, the soil thermal regime, and vegetation cover. We worked along 400-500 m transects at sites that represent the variety of ecotypes common in interior Alaska. Airborne LiDAR imagery was collected from May 9-11, 2014 with a spatial resolution of 0.25 m. During the winters from 2013-2019 snow pack depths have been made at roughly 1 m intervals along site transects using a snow depth datalogger coupled with a GPS. In late summer from 2013-2019 maximum seasonal thaw depths have been measured at 4 m intervals along each transect. Electrical resistivity tomography measurements were collected across the site transects. A variety of machine learning geospatial analysis approaches were also used to identify relationships between ecosystem characteristics, seasonal thaw, and permafrost soil and ice composition. Wintertime measurements show a clear relationship between vegetation cover and snow depth. Interception (and shallow snow) was evident in the birch and white spruce forests and where dense shrubs are present while the open tussock and intermittent shrub regions yield the greatest snow depths. Results from repeat seasonal thaw depth measurements also show a strong relationship with vegetation where mixed birch and spruce forest is associated with the deepest seasonal thaw. The tussock/shrub and spruce forest zones consistently exhibited the shallowest seasonal thaw. Roughly 60% of the seasonal thaw along the transects occurred by mid-July and downward movement of the thaw front had mostly ceased by late August with little additional thaw between August 20 and early October. Summer precipitation shows a relationship with seasonal thaw depth with the wettest summers associated with the deepest thaw. Results from this study identify clear relationships between ecotype, permafrost composition, and seasonal thaw dynamics that can help identify how and where permafrost degradation can be expected in a warmer future arctic.

Download Full-text

An Integrated Machine Learning Approach for Automatic Highway Extraction from Airborne LiDAR Data and Orthophotos

Laser Scanning Systems in Highway and Safety Assessment - Advances in Science, Technology & Innovation ◽

10.1007/978-3-030-10374-3_5 ◽

2019 ◽

pp. 61-76

Author(s):

Biswajeet Pradhan ◽

Maher Ibrahim Sameen

Keyword(s):

Machine Learning ◽

Airborne Lidar ◽

Lidar Data ◽

Learning Approach ◽

Machine Learning Approach ◽

Airborne Lidar Data

Download Full-text

Supplementary material to "Brief communication: Rapid machine learning-based extraction and measurement of ice wedge polygons in airborne lidar data"

10.5194/tc-2018-167-supplement ◽

2018 ◽

Author(s):

Charles J. Abolt ◽

Michael H. Young ◽

Adam A. Atchley ◽

Cathy J. Wilson

Keyword(s):

Machine Learning ◽

Airborne Lidar ◽

Lidar Data ◽

Ice Wedge ◽

Supplementary Material ◽

Airborne Lidar Data

Download Full-text

Using machine learning to generate high-resolution soil wetness maps for planning forest management

10.5194/egusphere-egu21-2578 ◽

2021 ◽

Author(s):

William Lidberg ◽

Johannes Larson ◽

siddhartho Paul ◽

Hjalmar Laudon ◽

Anneli Ågren

Keyword(s):

Machine Learning ◽

Soil Moisture ◽

High Resolution ◽

Field Data ◽

Airborne Lidar ◽

Aerial Photographs ◽

Riparian Zones ◽

Gradient Boosting ◽

Extreme Gradient Boosting ◽

Boreal Landscape

Open peatlands are a recognizable feature in the boreal landscape that are commonly mapped from aerial photographs. However, wet soils also occur on tree covered peatlands and in the riparian zones of forest streams and surrounding lakes. Comparisons between field data and available maps show that only 36 % of wet soils in the boreal landscape are marked on maps, making them difficult to manage. Wet soils have lower bearing capacity than dry soils and are more susceptible to soil disturbance from land-use management with heavy machinery. Topographical modelling of wet area indices has been suggested as a solution to this problem and high-resolution digital elevation models (DEM) derived from airborne LiDAR are becoming accessible in many countries. However, most of these topographical methods relies on the user to define appropriate threshold values in order to define wet areas. Soil textures, topography and climatic differences make any application difficult on a large scale. This complex landscape variability can be captured by utilizing machine learners that uses automated data mining methods to discover patterns in large data sets. By using soil moisture data from 20&#160;000 field plots from the National Forest Inventory of Sweden, we combined information from 24 indices and ancillary environmental features using a machine learning known as extreme gradient boosting. Extreme gradient boosting used the field data to learn how to classify soil moisture and delivered high performance compared to many traditional single algorithm methods. With this method we mapped soil moisture at 2 m spatial resolution across the Swedish forest landscape in five days using a workstation with 32 cores. This new map captured 79 % (kappa 0.69) of all wet soils compared to only 36 % (kappa 0.39) captured by current maps. In addition to capture open wetlands this new map also capture riparian zones and previously unmapped cryptic wetlands underneath the forest canopy. The new maps can, for example, be used to plan hydrologically adapted buffer zones, suggest machine free zones near streams and lakes in order to prevent rutting from forestry machines to reduce sediment, mercury and nutrient loads to downstream streams, lakes and sea.

Download Full-text

Classification of stones in coastal marine environments using random forest machine learning on topo-bathymetric LiDAR data

10.5194/egusphere-egu21-8254 ◽

2021 ◽

Author(s):

Signe Schilling Hansen ◽

Verner Brandbyge Ernstsen ◽

Mikkel Skovgaard Andersen ◽

Zyad Al-Hamdani ◽

Ramona Baran ◽

...

Keyword(s):

Machine Learning ◽

Ecosystem Services ◽

Random Forest ◽

Shallow Water ◽

Airborne Lidar ◽

Habitat Mapping ◽

Lidar Data ◽

Marine Environments ◽

Automated Method ◽

Coastal Marine

Stones on the seabed in coastal marine environments form an important hard substrate for macroalgae, and hence for coastal marine reefs. Such reef areas constitute important ecosystem services, e.g. storage of organic carbon in macroalgae or &#8220;blue carbon&#8221; as well as important habitats to fish for living, hiding and feeding. Information and knowledge about stone locations and geometry in coastal marine environments are often obtained as part of seabed habitat mapping. Usually, seabed habitat mapping is based on geophysical surveys using multibeam echo sounding along with side-scan sonar imaging in combination with biological ground-truthing. However, coastal areas are challenging to map with full spatial coverage due to the shallow water conditions. Furthermore, the research vessels often have too large drafts to sail in very shallow water close to the coastline. An alternative is to use airborne LiDAR technology. Topo-bathymetric LiDAR (green wavelength of 532&#160;nm) has made it possible to derive high-resolution data of the bathymetry in coastal zones (e.g. Andersen et al., 2017). This technology can cover the transition zone between land and water, and the time consumption for data acquisition is small compared to vessel borne methods. However, the processing of the data still requires manual decision steps, which makes it rather time consuming, and to some extent subjective.The aim of this study was to investigate the possibility of developing an automated method to classify stones from topo-bathymetric LiDAR data in coastal marine environments with shallow water (<6&#160;m). The R&#248;dsand lagoon in Denmark, where topo-bathymetric LiDAR data were acquired in 2015, was used as test. The classification was done using the random forest machine learning algorithm. The study resulted in the development of a nearly automated method to classify stones from topo-bathymetric LiDAR data. The classification accuracy was between 80 and 90% for the test site. The obtained knowledge about stone locations can provide important information about the ecosystem services and improved management of the coastal marine environment.&#160;Acknowledgement:This work is part of the project "ECOMAP - Baltic Sea environmental assessments by opto-acoustic remote sensing, mapping, and monitoring", supported by BONUS (Art 185), funded jointly by the EU and the Innovation Fund Denmark.&#160;ReferencesAndersen MS, Gergely A, Al-Hamdani Z, Steinbacher F, Larsen LR, Ernstsen VB (2017). Processing and performance of topobathymetric lidar data for geomorphometric and morphological classification in a high-energy tidal environment. Hydrology and Earth System Sciences, 21: 43-63, DOI:&#160;10.5194/hess-21-43-2017.

Download Full-text