scholarly journals Uncertainty and Overfitting in Fluvial Landform Classification Using Laser Scanned Data and Machine Learning: A Comparison of Pixel and Object-Based Approaches

2020 ◽  
Vol 12 (21) ◽  
pp. 3652 ◽  
Author(s):  
Zsuzsanna Csatáriné Szabó ◽  
Tomáš Mikita ◽  
Gábor Négyesi ◽  
Orsolya Gyöngyi Varga ◽  
Péter Burai ◽  
...  
Keyword(s):  
Laser Scanning ◽  
Model Performance ◽  
Digital Terrain Model ◽  
Recursive Feature Elimination ◽  
Random Forest Classification ◽  
Terrain Model ◽  
Landform Classification ◽  
Forest Classification ◽  
Object Based ◽  
Geomorphological Map

Floodplains are valuable scenes of water management and nature conservation. A better understanding of their geomorphological characteristic helps to understand the main processes involved. We performed a classification of floodplain forms in a naturally developed area in Hungary using a Digital Terrain Model (DTM) of aerial laser scanning. We derived 60 geomorphometric variables from the DTM and prepared a geomorphological map of 265 forms (crevasse channels, point bars, swales, levees). Random Forest classification was conducted with Recursive Feature Elimination (RFE) on the objects (mean pixel values by forms) and on the pixels of the variables. We also evaluated the classification probabilities (CP), the spatial uncertainties (SU), and the overfitting in the function of the number of the variables. We found that the object-based method had a better performance (95%) than the pixel-based method (78%). RFE helped to identify the most important 13–20 variables, maintaining the high model performance and reducing the overfitting. However, CP and SU were not efficient measures of classification accuracy as they were not in accordance with the class level accuracy metric. Our results help to understand classification results and the specific limits of laser scanned DTMs. This methodology can be useful in geomorphologic mapping.

scholarly journals AN OBJECT-BASED METHOD FOR CHINESE LANDFORM TYPES CLASSIFICATION

2016 ◽  
Vol XLI-B7 ◽  
pp. 213-217 ◽  
Author(s):  
Hu Ding ◽  
Fei Tao ◽  
Wufan Zhao ◽  
Jiaming Na ◽  
Guo’an Tang
Keyword(s):  
Random Forest ◽  
Classification Tree ◽  
Unsupervised Classification ◽  
Entropy Method ◽  
Random Forest Classification ◽  
Landform Classification ◽  
Forest Classification ◽  
Object Based ◽  
Importance Analysis ◽  
Occurrence Matrix

Landform classification is a necessary task for various fields of landscape and regional planning, for example for landscape evaluation, erosion studies, hazard prediction, et al. This study proposes an improved object-based classification for Chinese landform types using the factor importance analysis of random forest and the gray-level co-occurrence matrix (GLCM). In this research, based on 1km DEM of China, the combination of the terrain factors extracted from DEM are selected by correlation analysis and Sheffield's entropy method. Random forest classification tree is applied to evaluate the importance of the terrain factors, which are used as multi-scale segmentation thresholds. Then the GLCM is conducted for the knowledge base of classification. The classification result was checked by using the 1:4,000,000 Chinese Geomorphological Map as reference. And the overall classification accuracy of the proposed method is 5.7% higher than ISODATA unsupervised classification, and 15.7% higher than the traditional object-based classification method.

scholarly journals AN OBJECT-BASED METHOD FOR CHINESE LANDFORM TYPES CLASSIFICATION

2016 ◽  
Vol XLI-B7 ◽  
pp. 213-217
Author(s):  
Hu Ding ◽  
Fei Tao ◽  
Wufan Zhao ◽  
Jiaming Na ◽  
Guo’an Tang
Keyword(s):  
Random Forest ◽  
Classification Tree ◽  
Unsupervised Classification ◽  
Entropy Method ◽  
Random Forest Classification ◽  
Landform Classification ◽  
Forest Classification ◽  
Object Based ◽  
Importance Analysis ◽  
Occurrence Matrix

Landform classification is a necessary task for various fields of landscape and regional planning, for example for landscape evaluation, erosion studies, hazard prediction, et al. This study proposes an improved object-based classification for Chinese landform types using the factor importance analysis of random forest and the gray-level co-occurrence matrix (GLCM). In this research, based on 1km DEM of China, the combination of the terrain factors extracted from DEM are selected by correlation analysis and Sheffield's entropy method. Random forest classification tree is applied to evaluate the importance of the terrain factors, which are used as multi-scale segmentation thresholds. Then the GLCM is conducted for the knowledge base of classification. The classification result was checked by using the 1:4,000,000 Chinese Geomorphological Map as reference. And the overall classification accuracy of the proposed method is 5.7% higher than ISODATA unsupervised classification, and 15.7% higher than the traditional object-based classification method.

scholarly journals DTM-based analysis of the spatial distribution of topolineaments

2020 ◽  
Vol 12 (1) ◽  
pp. 1185-1199
Author(s):  
Mirosław Kamiński
Keyword(s):  
Laser Scanning ◽  
Digital Terrain Model ◽  
Research Area ◽  
Tectonic Structures ◽  
Empirical Bayesian ◽  
Scanning Method ◽  
Empirical Bayesian Kriging ◽  
Terrain Model ◽  
Digital Terrain ◽  
Airborne Laser

AbstractThe research area is located on the boundary between two Paleozoic structural units: the Radom–Kraśnik Block and the Mazovian–Lublin Basin in the southeastern Poland. The tectonic structures are separated by the Ursynów–Kazimierz Dolny fault zone. The digital terrain model obtained by the ALS (Airborne Laser Scanning) method was used. Classification and filtration of an elevation point cloud were performed. Then, from the elevation points representing only surfaces, a digital terrain model was generated. The model was used to visually interpret the course of topolineaments and their automatic extraction from DTM. Two topolineament systems, trending NE–SW and NW–SE, were interpreted. Using the kernel density algorithm, topolineament density models were generated. Using the Empirical Bayesian Kriging, a thickness model of quaternary deposits was generated. A relationship was observed between the course of topolineaments and the distribution and thickness of Quaternary formations. The topolineaments were compared with fault directions marked on tectonic maps of the Paleozoic and Mesozoic. Data validation showed consistency between topolineaments and tectonic faults. The obtained results are encouraging for further research.

scholarly journals TECHNIQUE OF AUTOMATIC MOBILE LASER SCANNING DATA FILTERING

2021 ◽  
Vol 26 (3) ◽  
pp. 5-19
Author(s):  
Maxim A. Altyntsev ◽  
◽  
Hamid Majid Saber Karkokli ◽  
Keyword(s):  
Laser Scanning ◽  
Single Point ◽  
Laser Scanner ◽  
Digital Terrain Model ◽  
Weather Conditions ◽  
Automatic Generation ◽  
Data Filtering ◽  
Terrain Model ◽  
Digital Terrain ◽  
Filtration Technique

The result of laser scanning is an array of laser points. The generation of a single point cloud in a given coordinate system is carried out during the registration process at the stage of preliminary field data processing. At this stage it is also often necessary to filter the data. Laser points with an erroneous position are eliminated during the data filleting. The number of erroneous laser points is determined by the of the laser scanner characteristics, surveyed area peculiarities and weather conditions. The devel-opment of methods and algorithms for filtering laser scanning data is carried out based on the analysis of the laser point spatial position and a certain set of additional characteristics, such as intensity value, echo signal, color value. The technique of mobile laser scanning data filtering for the territory of the road passing among the forest and close to individual industrial facilities and building. The main goal of the proposed filtration technique is to obtain data for automatic generation of an accurate digital terrain model. The filtration technique was developed for data acquired under the least favorable con-ditions – in wet weather. Accuracy estimation of generating digital terrain model based on filtered data was carried out.

scholarly journals Automating the creation of channel cross section data from aerial laser scanning and hydrological surveying for modeling flood events / Automatická tvorba geometrie vodních toků na základě syntézy dat z leteckého laserového skenování a hydrologického měření pro modelování povodňových událostí. j. hydrol., hydromech., 60, 2012, 4; 25 lit., 8 obr., 5 tab.

2012 ◽  
Vol 60 (4) ◽  
pp. 227-241 ◽  
Author(s):  
Radek Roub ◽  
Tomáš Hejduk ◽  
Pavel Novák
Keyword(s):  
Flow Rate ◽  
Hydrodynamic Model ◽  
Input Data ◽  
Laser Scanning ◽  
Digital Terrain Model ◽  
Flood Events ◽  
Hydrodynamic Models ◽  
Terrain Model ◽  
Aerial Laser Scanning ◽  
Digital Terrain

Knowing the extent of inundation areas for individual N-year flood events, the specific flood scenarios, and having an idea about the depths and velocities in the longitudinal or transverse water course profile provided by hydrodynamic models is of key importance for protecting peoples’ lives and mitigating damage to property. Input data for creating the watercourse computational geometry are crucial for hydrodynamic models. Requirements for input data vary with respect to the hydrodynamic model used. One-dimensional (1D) hydrodynamic models in which the computing track is formed by cross-sectional profiles of the channel are characterized by lower requirements for input data. In two-dimensional (2D) hydrodynamic models, a digital terrain model is needed for the entire area studied. Financial requirements of the project increase with regard to the input data and the model used. The increase is mainly due to the high cost of the geodetic surveying of the stream channel. The paper aims at a verification and presentation of the suitability of using hydrological measurements in developing a schematization (geometry) of water courses based on topographic data gained from aerial laser scanning provided by the Czech Office for Surveying, Mapping and Cadastre. Taking into account the hydrological measurement during the schematization of the water course into the hydrodynamic model consists in the derivation of flow rate achieved at the time of data acquisition using the method of aerial laser scanning by means of hydrological analogy and in using the established flow rate values as a basis for deepening of the digital terrain model from aerial laser scanning data. Thus, the given principle helps to capture precisely the remaining part of the channel profile which is not reflected in the digital terrain model prepared by the method of aerial laser scanning and fully correct geometry is achieved for the hydrodynamic model.

scholarly journals Threat of Pollution Hotspots Reworking in River Systems: Case Study of the Ploučnice River (Czech Republic)

2019 ◽  
Vol 8 (1) ◽  
pp. 37 ◽  
Author(s):  
Jitka Elznicová ◽  
Tomáš Matys Grygar ◽  
Jan Popelka ◽  
Martin Sikora ◽  
Petr Novák ◽  
...  
Keyword(s):  
Laser Scanning ◽  
Digital Terrain Model ◽  
Terrain Model ◽  
Non Invasive ◽  
Digital Terrain ◽  
Quality Of Water ◽  
Lateral Channel ◽  
Floodplain Development ◽  
Channel Deposits

As fluvial pollution may endanger the quality of water and solids transported by rivers, mapping and evaluation of historically polluted fluvial sediments is an urgent topic. The Ploučnice River and its floodplain were polluted by local uranium mining from 1971–1989. We have studied this river since 2013 using a combination of diverse methods, including geoinformatics, to identify pollution hotspots in floodplains and to evaluate the potential for future reworking. Archival information on pollution history and past flooding was collected to understand floodplain dynamics and pollution heterogeneity. Subsequently, a digital terrain model based on laser scanning data and data analysis were used to identify the sites with river channel shifts. Finally, non-invasive geochemical mapping was employed, using portable X-ray fluorescence and gamma spectrometers. The resulting datasets were processed with geostatistical tools. One of the main outputs of the study was a detailed map of pollution distribution in the floodplain. The results showed a relationship between polluted sediment deposition, past channel shifts and floodplain development. We found that increased concentration of pollution occurred mainly in the cut-off meanders and lateral channel deposits from the mining period, the latter in danger of reworking (reconnecting to the river) in the coming decades.

scholarly journals UAV-Based Terrain Modeling under Vegetation in the Chinese Loess Plateau: A Deep Learning and Terrain Correction Ensemble Framework

2020 ◽  
Vol 12 (20) ◽  
pp. 3318 ◽  
Author(s):  
Jiaming Na ◽  
Kaikai Xue ◽  
Liyang Xiong ◽  
Guoan Tang ◽  
Hu Ding ◽  
...  
Keyword(s):  
Deep Learning ◽  
Loess Plateau ◽  
Laser Scanning ◽  
Digital Terrain Model ◽  
Terrain Correction ◽  
Chinese Loess Plateau ◽  
Chinese Loess ◽  
Terrain Model ◽  
Vegetation Height ◽  
Digital Terrain

Accurate topographic mapping is a critical task for various environmental applications because elevation affects hydrodynamics and vegetation distributions. UAV photogrammetry is popular in terrain modelling because of its lower cost compared to laser scanning. However, this method is restricted in vegetation area with a complex terrain, due to reduced ground visibility and lack of robust and automatic filtering algorithms. To solve this problem, this work proposed an ensemble method of deep learning and terrain correction. First, image matching point cloud was generated by UAV photogrammetry. Second, vegetation points were identified based on U-net deep learning network. After that, ground elevation was corrected by estimating vegetation height to generate the digital terrain model (DTM). Two scenarios, namely, discrete and continuous vegetation areas were considered. The vegetation points in the discrete area were directly removed and then interpolated, and terrain correction was applied for the points in the continuous areas. Case studies were conducted in three different landforms in the loess plateau of China, and accuracy assessment indicated that the overall accuracy of vegetation detection was 95.0%, and the MSE (Mean Square Error) of final DTM (Digital Terrain Model) was 0.024 m.

scholarly journals A novel automated method for the improvement of photogrammetric DTM accuracy in forests

2018 ◽  
Vol 142 (11-12) ◽  
pp. 576-577 ◽  
Author(s):  
Mateo Gašparović ◽  
Ivan Balenović ◽  
Ante Seletković ◽  
Anita Simic Milas
Keyword(s):  
Quercus Robur ◽  
Laser Scanning ◽  
Digital Terrain Model ◽  
Light Detection And Ranging ◽  
Mean Square ◽  
Terrain Model ◽  
Digital Terrain ◽  
Automated Method ◽  
Airborne Laser ◽  
Fraxinus Angustifolia

Digitalni model reljefa (DTM, engl. Digital Terrain Model) ima široku i važnu primjenu u mnogim djelatnostima, uključujući i šumarstvo. Međutim, precizno modeliranje terena, odnosno izrada DTM-a u šumama, bilo korištenjem terenskih metoda ili metoda daljinskih istraživanja, izazovan je i vrlo zahtjevan zadatak. U većini razvijenih zemalja svijeta, zračno lasersko skeniranje (ALS, engl. Airborne Laser Scanning) bazirano na LiDAR (engl. Light Detection and Ranging) tehnologiji trenutno predstavlja glavnu metodu za izradu DTM-a. Uslijed mogućnosti laserskog zračenja da penetrira kroz krošnje drveća, LiDAR tehnologija se pokazala kao efektivna i brza metoda za izradu DTM-a u šumskim područjima s vrlo velikom točnošću. Međutim, u mnogim zemljama svijeta, uključujući i Hrvatsku, zračno lasersko skeniranje nije u potpunosti provedeno, tj. samo su manji dijelovi zemlje pokriveni s podacima zračnog laserskog skeniranja. U tim slučajevima, DTM temeljen na stereo-fotogrametrijskoj izmjeri aerosnimaka potpomognut s terenskim podacima najčešće predstavlja glavni izvor informacija za izradu DTM-a. Poznato je da tako izrađen DTM u šumskim predjelima ima manju točnost od DTM-a dobivenog na temelju zračnog laserskog skeniranja zbog pokrivenosti terena vegetacijom. Također, u okviru nedavno provedenog istraživanja (Balenović i dr., 2018) utvrđeno je da takvi službeni fotogrametrijski digitalni podaci terena u šumskim predjelima sadrže određen broj tzv. grubih grešaka, koje mogu značajno utjecati na točnost izrađenog DTM-a. Nakon vizualnog detektiranja i manualnog uklanjanja tih pogrešaka, Balenović i dr. (2018) utvrdili su značajno poboljšanje točnosti fotogrametrijskog DTM-a. Stoga je glavni cilj ovoga rada razviti automatsku metodu za detekciju i eliminaciju vertikalnih pogrešaka u fotogrametrijskim digitalnim podacima terena te na taj način poboljšati točnost fotogrametrijskog DTM-a u nizinskim šumskim područjima Hrvatske. Ideja je razviti brzu, jednostavnu i učinkovitu metodu koja će biti primjenjiva i za druga šumska područja sličnih karakteristika, a za koja ne postoje DTM dobiven zračnim laserskim skeniranjem. Istraživanje je provedeno u nizinskim šumama na području gospodarske jedinice Jastrebarski lugovi, u neposrednoj blizini Jastrebarskog (Slika 1). Istraživanjem je obuhvaćena površina od 2.005,74 ha, na kojoj su u najvećoj mjeri zastupljene jednodobne sastojine hrasta lužnjaka (Quercus robur L.), a u ma­njoj mjeri jednodobne sastojine poljskog jasena (Fraxinus angustifolia L.) te jednodobne sastojine običnoga graba (Carpinus betulus L.). Nadmorska visina područja istraživanja kreće se u rasponu od 105 do 121 m. Fotogrametrijski DTM (DTM<sub>PHM</sub>) je izrađen iz digitalnih vektorskih podataka terena (prijelomnice, linije oblika, markantne točke terena i pravokutne mreže visinskih točaka) nabavljenih iz Državne geodetske uprave (Slika 2). Ti podaci predstavljaju nacionalni standard i jedini su dostupni podaci za izradu DTM-a u Hrvatskoj. Detaljan opis vektorskih podataka dan je u radu Balenović i dr. (2018). Prvo je iz digitalnih terenskih podataka izrađena nepravilna mreža trokuta, koja je potom linearnom interpolacijom pretvorena u rasterski DTM<sub>PHM</sub> prostorne rezolucije (veličine piksela) 0,5 m. Automatska metoda za detekciju i eliminaciju vertikalnih pogrešaka fotogrametrijskog DTM-a u nizinskim šumskim područjima razvijena je u slobodnom programskom paketu Grass GIS (Slika 3). Kombinacijom vrijednosti nagiba i tangencijalne zakrivljenosti terena rasterskog DTM<sub>PHM</sub> (Slika 4), automatskom metodom su detektirane 91 grube greške (engl. outliers). Drugim riječima, utvrđeno je da 91 točkasti vektorski objekt pogrešno prikazuje stvarnu visinu terena. Navedeni broj čini 3,2 % od ukupnog broja točkastih objekata korištenih za izradu DTM<sub>PHM</sub>-a. Nakon eliminacije detektiranih pogrešaka izrađen je novi, korigirani fotogrametrijski DTM (DTM<sub>PHMc</sub>). Za ocjenu vertikalne točnosti izvornog (DTM<sub>PHM</sub>) i korigiranog DTM-a (DTM<sub>PHMc</sub>) korišten je visoko precizni DTM dobiven zračnim laserskim skeniranjem (DTM<sub>LiD</sub>). U tu svrhu su izrađeni rasteri razlika između DTM<sub>PHM </sub>i DTM<sub>LiD</sub>, te između DTM<sub>PHMc </sub>i DTM<sub>LiD</sub>. Kako je preliminarnom analizom utvrđeno da vertikalne razlike između DTM<sub>PHM </sub>i DTM<sub>LiD</sub> nisu normalno distribuirane (Slika 5), za ocjenu točnosti su uz normalne mjere točnosti korištene i tzv. robusne mjere točnosti (Tablica 2). Dobiveni rezultati ukazuju na poboljšanje vertikalne točnosti fotogrametrijskog DTM-a primjenom razvijene automatske metode. To je posebice uočljivo na podpodručjima 2 i 3 (Slika 6 i 7) u kojima se nakon uklanjanja detektiranih grešaka, korijen srednje kvadratne pogreške (RMSE, engl. root mean square error) smanjio za 8 % odnosno 50 % (Tablica 2). Na temelju dobivenih rezultata i usporedbe s DTM<sub>LiD</sub>, može se zaključiti da predložena metoda uspješno detektira i eliminira vertikalne pogreške fotogrametrijskog DTM-a u nizinskim šumskim područjima, te slijedom toga poboljšava njegovu vertikalnu točnost.

scholarly journals Reduction Method for Mobile Laser Scanning Data

2018 ◽  
Vol 7 (7) ◽  
pp. 285 ◽  
Author(s):  
Wioleta Błaszczak-Bąk ◽  
Zoltan Koppanyi ◽  
Charles Toth
Keyword(s):  
Data Reduction ◽  
Point Cloud ◽  
Laser Scanning ◽  
Digital Terrain Model ◽  
New Approach ◽  
Terrain Model ◽  
Digital Terrain ◽  
Airborne Laser ◽  
Short Time ◽  
Mls Method

Mobile Laser Scanning (MLS) technology acquires a huge volume of data in a very short time. In many cases, it is reasonable to reduce the size of the dataset with eliminating points in such a way that the datasets, after reduction, meet specific optimization criteria. Various methods exist to decrease the size of point cloud, such as raw data reduction, Digital Terrain Model (DTM) generalization or generation of regular grid. These methods have been successfully applied on data captured from Airborne Laser Scanning (ALS) and Terrestrial Laser Scanning (TLS), however, they have not been fully analyzed on data captured by an MLS system. The paper presents our new approach, called the Optimum Single MLS Dataset method (OptD-single-MLS), which is an algorithm for MLS data reduction. The tests were carried out in two variants: (1) for raw sensory measurements and (2) for a georeferenced 3D point cloud. We found that the OptD-single-MLS method provides a good solution in both variants; therefore, the choice of the reduction variant depends only on the user.

Sign in / Sign up

Export Citation Format

Share Document