Centre dedocumentation
scientifique

Comparison of neural networks and k-nearest neighbors methods in forest stand variable estimation using airborne laser data / Andras Balazs in ISPRS Open Journal of Photogrammetry and Remote Sensing, vol 4 (April 2022)  

Public [article]  inISPRS Open Journal of Photogrammetry and Remote Sensing > vol 4 (April 2022) . - n° 100012 Titre : Comparison of neural networks and k-nearest neighbors methods in forest stand variable estimation using airborne laser data Type de document : Article/Communication Auteurs : Andras Balazs, Auteur ; Eero Liski, Auteur ; Sakari Tuominen, Auteur Année de publication : 2022 Article en page(s) : n° 100012 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes IGN] algorithme génétique [Termes IGN] bois sur pied [Termes IGN] classification barycentrique [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] covariance [Termes IGN] diamètre à hauteur de poitrine [Termes IGN] données lidar [Termes IGN] données localisées 3D [Termes IGN] Finlande [Termes IGN] hauteur des arbres [Termes IGN] inventaire forestier étranger (données) [Termes IGN] peuplement forestier [Termes IGN] réseau neuronal artificiel [Termes IGN] semis de points [Termes IGN] volume en bois Résumé : (auteur) In the remote sensing of forests, point cloud data from airborne laser scanning contains high-value information for predicting the volume of growing stock and the size of trees. At the same time, laser scanning data allows a very high number of potential features that can be extracted from the point cloud data for predicting the forest variables. In some methods, the features are first extracted by user-defined algorithms and the best features are selected based on supervised learning, whereas both tasks can be carried out automatically by deep learning methods typically based on deep neural networks. In this study we tested k-nearest neighbor method combined with genetic algorithm (k-NN), artificial neural network (ANN), 2-dimensional convolutional neural network (2D-CNN) and 3-dimensional CNN (3D-CNN) for estimating the following forest variables: volume of growing stock, stand mean height and mean diameter. The results indicate that there were no major differences in the accuracy of the tested methods, but the ANN and 3D-CNN generally resulted in the lowest RMSE values for the predicted forest variables and the highest R2 values between the predicted and observed forest variables. The lowest RMSE scores were 20.3% (3D-CNN), 6.4% (3D-CNN) and 11.2% (ANN) and the highest R2 results 0.90 (3D-CNN), 0.95 (3D-CNN) and 0.85 (ANN) for volume of growing stock, stand mean height and mean diameter, respectively. Covariances of all response variable combinations and all predictions methods were lower than corresponding covariances of the field observations. ANN predictions had the highest covariances for mean height vs. mean diameter and total growing stock vs. mean diameter combinations and 3D-CNN for mean height vs. total growing stock. CNNs have distinct theoretical advantage over the other methods in complex recognition or classification tasks, but the utilization of their full potential may possibly require higher point density clouds than applied here. Thus, the relatively low density of the point clouds data may have been a contributing factor to the somewhat inconclusive ranking of the methods in this study. The input data and computer codes are available at: https://github.com/balazsan/ALS_NNs. Numéro de notice : A2022-265 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.ophoto.2022.100012 Date de publication en ligne : 12/03/2022 En ligne : https://doi.org/10.1016/j.ophoto.2022.100012 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100263 [article]

Economic losses caused by tree species proportions and site type errors in forest management planning / Arto Haara in Silva fennica, vol 53 n° 2 (2019)  

Public [article]  inSilva fennica > vol 53 n° 2 (2019) Titre : Economic losses caused by tree species proportions and site type errors in forest management planning Type de document : Article/Communication Auteurs : Arto Haara, Auteur ; Annika S. Kangas, Auteur ; Sakari Tuominen, Auteur Année de publication : 2019 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] coupe (sylviculture) [Termes IGN] données lidar [Termes IGN] données localisées 3D [Termes IGN] éclaircie (sylviculture) [Termes IGN] erreur [Termes IGN] Finlande [Termes IGN] identification de plantes [Termes IGN] image 3D [Termes IGN] image aérienne [Termes IGN] image spatiale [Termes IGN] incertitude des données [Termes IGN] inventaire forestier étranger (données) [Vedettes matières IGN] Economie forestière Résumé : (auteur) The aim of this study was to estimate economic losses, which are caused by forest inventory errors of tree species proportions and site types. Our study data consisted of ground truth data and four sets of erroneous tree species proportions. They reflect the accuracy of tree species proportions in four remote sensing data sets, namely 1) airborne laser scanning (ALS) with 2D aerial image, 2) 2D aerial image, 3) 3D and 2D aerial image data together and 4) satellite data. Furthermore, our study data consisted of one simulated site type data set. We used the erroneous tree species proportions to optimise the timing of forest harvests and compared that to the true optimum obtained with ground truth data. According to the results, the mean losses of Net Present Value (NPV) because of erroneous tree species proportions at an interest rate of 3% varied from 124.4 € ha–1 to 167.7 € ha–1. The smallest losses were observed using tree species proportions predicted using ALS data and largest using satellite data. In those stands, respectively, in which tree species proportion errors actually caused economic losses, they were 468 € ha–1 on average with tree species proportions based on ALS data. In turn, site type errors caused only small losses. Based on this study, accurate tree species identification seems to be very important with respect to operational forest inventory. Numéro de notice : A2019-378 Affiliation des auteurs : non IGN Thématique : FORET Nature : Article DOI : 10.14214/sf.10089 Date de publication en ligne : 17/06/2019 En ligne : https://doi.org/10.14214/sf.10089 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=93444 [article]

Hyperspectral UAV-imagery and photogrammetric canopy height model in estimating forest stand variables / Sakari Tuominen in Silva fennica, vol 51 n° 5 (2017)  

Public [article]  inSilva fennica > vol 51 n° 5 (2017) Titre : Hyperspectral UAV-imagery and photogrammetric canopy height model in estimating forest stand variables Type de document : Article/Communication Auteurs : Sakari Tuominen, Auteur ; Andras Balazs, Auteur ; Eija Honkavaara, Auteur ; et al., Auteur Année de publication : 2017 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] classification barycentrique [Termes IGN] diamètre des arbres [Termes IGN] étalonnage radiométrique [Termes IGN] hauteur des arbres [Termes IGN] image aérienne [Termes IGN] image captée par drone [Termes IGN] image hyperspectrale [Termes IGN] image RVB [Termes IGN] inventaire forestier (techniques et méthodes) [Termes IGN] modèle numérique de surface de la canopée [Termes IGN] peuplement forestier [Termes IGN] photogrammétrie numérique [Termes IGN] volume en bois [Vedettes matières IGN] Inventaire forestier Résumé : (auteur) Remote sensing using unmanned aerial vehicle (UAV) -borne sensors is currently a highly interesting approach for the estimation of forest characteristics. 3D remote sensing data from airborne laser scanning or digital stereo photogrammetry enable highly accurate estimation of forest variables related to the volume of growing stock and dimension of the trees, whereas recognition of tree species dominance and proportion of different tree species has been a major complication in remote sensing-based estimation of stand variables. In this study, the use of UAV-borne hyperspectral imagery was examined in combination with a high-resolution photogrammetric canopy height model in estimating forest variables of 298 sample plots. Data were captured from eleven separate test sites under weather conditions varying from sunny to cloudy and partially cloudy. Both calibrated hyperspectral reflectance images and uncalibrated imagery were tested in combination with a canopy height model based on RGB camera imagery using the k-nearest neighbour estimation method. The results indicate that this data combination allows accurate estimation of stand volume, mean height and diameter: the best relative RMSE values for those variables were 22.7%, 7.4% and 14.7%, respectively. In estimating volume and dimension-related variables, the use of a calibrated image mosaic did not bring significant improvement in the results. In estimating the volumes of individual tree species, the use of calibrated hyperspectral imagery generally brought marked improvement in the estimation accuracy; the best relative RMSE values for the volumes for pine, spruce, larch and broadleaved trees were 34.5%, 57.2%, 45.7% and 42.0%, respectively. Numéro de notice : A2017-645 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article DOI : 10.14214/sf.7721 En ligne : https://doi.org/10.14214/sf.7721 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=87000 [article]

Improving Finnish multi-source national forest inventory by 3D aerial imaging / Sakari Tuominen in Silva fennica, vol 51 n° 4 (2017)

Public [article]  inSilva fennica > vol 51 n° 4 (2017) Titre : Improving Finnish multi-source national forest inventory by 3D aerial imaging Type de document : Article/Communication Auteurs : Sakari Tuominen, Auteur ; Timo P Pitkänen, Auteur ; Andras Balazs, Auteur ; et al., Auteur Année de publication : 2017 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] classification barycentrique [Termes IGN] corrélation automatique de points homologues [Termes IGN] distribution spatiale [Termes IGN] Finlande [Termes IGN] image aérienne [Termes IGN] image satellite [Termes IGN] inventaire forestier (techniques et méthodes) [Termes IGN] inventaire forestier étranger (données) [Termes IGN] modèle numérique de surface de la canopée [Termes IGN] photogrammétrie numérique [Termes IGN] placette d'échantillonnage [Vedettes matières IGN] Inventaire forestier Résumé : (auteur) Optical 2D remote sensing techniques such as aerial photographing and satellite imaging have been used in forest inventory for a long time. During the last 15 years, airborne laser scanning (ALS) has been adopted in many countries for the estimation of forest attributes at stand and sub-stand levels. Compared to optical remote sensing data sources, ALS data are particularly well-suited for the estimation of forest attributes related to the physical dimensions of trees due to its 3D information. Similar to ALS, it is possible to derive a 3D forest canopy model based on aerial imagery using digital aerial photogrammetry. In this study, we compared the accuracy and spatial characteristics of 2D satellite and aerial imagery as well as 3D ALS and photogrammetric remote sensing data in the estimation of forest inventory variables using k-NN imputation and 2469 National Forest Inventory (NFI) sample plots in a study area covering approximately 5800 km2. Both 2D data were very close to each other in terms of accuracy, as were both the 3D materials. On the other hand, the difference between the 2D and 3D materials was very clear. The 3D data produce a map where the hotspots of volume, for instance, are much clearer than with 2D remote sensing imagery. The spatial correlation in the map produced with 2D data shows a lower short-range correlation, but the correlations approach the same level after 200 meters. The difference may be of importance, for instance, when analyzing the efficiency of different sampling designs and when estimating harvesting potential. Numéro de notice : A2017-646 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=87003 [article]

Estimation of local forest attributes, utilizing two-phase sampling and auxiliary data / Sakari Tuominen (2007)  

Public Titre : Estimation of local forest attributes, utilizing two-phase sampling and auxiliary data Type de document : Thèse/HDR Auteurs : Sakari Tuominen, Auteur ; Simo Poso, Directeur de thèse Editeur : Helsinki [Finlande] : Faculty of Agriculture and Forestry of the University of Helsinki Année de publication : 2007 Autre Editeur : Helsinki [Finland] : The Finnish Society of Forest Science Importance : 46 p. Format : 21 x 30 cm ISBN/ISSN/EAN : 978-951-651-170-5 Note générale : bibliographie Academic dissertation to be presented, with the permission of the Faculty of Agriculture and Forestry of University of Helsinki, for public criticism in Auditorium 2, Forest Sciences' Building, Latokartanonkaari 9, Helsinki Langues : Anglais (eng) Descripteur : [Termes IGN] analyse comparative [Termes IGN] base de données forestières [Termes IGN] échantillonnage [Termes IGN] extraction de données [Termes IGN] Finlande [Termes IGN] forêt [Termes IGN] fusion de données multisource [Termes IGN] gestion forestière [Termes IGN] image à très haute résolution [Termes IGN] image aérienne [Termes IGN] image spectrale [Termes IGN] inventaire forestier (techniques et méthodes) [Termes IGN] inventaire forestier étranger (données) [Termes IGN] placette d'échantillonnage [Vedettes matières IGN] Inventaire forestier Résumé : (auteur) This thesis examines the feasibility of a forest inventory method based on two-phase sampling in estimating forest attributes at the stand or substand levels for forest management purposes. The method is based on multi-source forest inventory combining auxiliary data consisting of remote sensing imagery or other geographic information and field measurements. Auxiliary data are utilized as first-phase data for covering all inventory units. Various methods were examined for improving the accuracy of the forest estimates. Pre-processing of auxiliary data in the form of correcting the spectral properties of aerial imagery was examined (I), as was the selection of aerial image features for estimating forest attributes (II). Various spatial units were compared for extracting image features in a remote sensing aided forest inventory utilizing very high resolution imagery (III). A number of data sources were combined and different weighting procedures were tested in estimating forest attributes (IV, V). Correction of the spectral properties of aerial images proved to be a straightforward and advantageous method for improving the correlation between the image features and the measured forest attributes. Testing different image features that can be extracted from aerial photographs (and other very high resolution images) showed that the images contain a wealth of relevant information that can be extracted only by utilizing the spatial organization of the image pixel values. Furthermore, careful selection of image features for the inventory task generally gives better results than inputting all extractable features to the estimation procedure. When the spatial units for extracting very high resolution image features were examined, an approach based on image segmentation generally showed advantages compared with a traditional sample plot-based approach. Combining several data sources resulted in more accurate estimates than any of the individual data sources alone. The best combined estimate can be derived by weighting the estimates produced by the individual data sources by the inverse values of their mean square errors. Despite the fact that the plot-level estimation accuracy in two-phase sampling inventory can be improved in many ways, the accuracy of forest estimates based mainly on single-view satellite and aerial imagery is a relatively poor basis for making stand-level management decisions. Note de contenu : Introduction 1 - Two-phase sampling in forest inventory 2 - Remote sensing in forest inventory 3 - Objectives of the thesis and substudies 4 - Materials 5 - Results 6 - Discussion Numéro de notice : 21697 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Thèse étrangère Note de thèse : Thesis : Forestry : Helsinki : 2007 En ligne : https://helda.helsinki.fi/handle/10138/20667 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=90940

Performance of different spectral and textural photograph features in multi-source forest inventory / Sakari Tuominen in Remote sensing of environment, vol 94 n° 2 (30/01/2005)  

Permalink