Classification of forest development stages from national low-density lidar datasets: a comparison of machine learning methods


  • R. Valbuena University of Eastern Finland
  • M. Maltamo University of Eastern Finland
  • P. Packalen University of Eastern Finland



airborne laser scanning, discriminant analysis, maximum likelihood, minimum volume ellipsoid, naïve Bayes, support vector machine, artificial neural networks, random forest, nearest neighbour.


The area-based method has become a widespread approach in airborne laser scanning (ALS), being mainly employed for the estimation of continuous variables describing forest attributes: biomass, volume, density, etc. However, to date, classification methods based on machine learning, which are fairly common in other remote sensing fields, such as land use / land cover classification using multispectral sensors, have been largely overseen in forestry applications of ALS. In this article, we wish to draw the attention on statistical methods predicting discrete responses, for supervised classification of ALS datasets. A wide spectrum of approaches are reviewed: discriminant analysis (DA) using various classifiers –maximum likelihood, minimum volume ellipsoid, naïve Bayes–, support vector machine (SVM), artificial neural networks (ANN), random forest (RF) and nearest neighbour (NN) methods. They are compared in the context of a classification of forest areas into development classes (DC) used in practical silvicultural management in Finland, using their low-density national ALS dataset. We observed that RF and NN had the most balanced error matrices, with cross-validated predictions which were mainly unbiased for all DCs. Although overall accuracies were higher for SVM and ANN, their results were very dissimilar across DCs, and they can therefore be only advantageous if certain DCs are targeted. DA methods underperformed in comparison to other alternatives, and were only advantageous for the detection of seedling stands. These results show that, besides the well demonstrated capacity of ALS for quantifying forest stocks, there is a great deal of potential for predicting categorical variables in general, and forest types in particular. In conclusion, we consider that the presented methodology shall also be adapted to the type of forest classes that can be relevant to Mediterranean ecosystems, opening a range of possibilities for future research, in which ALS may show great predictive potential.


Download data is not yet available.

Author Biographies

R. Valbuena, University of Eastern Finland

School of Forest Sciences

M. Maltamo, University of Eastern Finland

School of Forest Sciences

P. Packalen, University of Eastern Finland

School of Forest Sciences


Atkinson, P.M., Tatnall, A.R.L. 1997. Introduction: Neural networks in remote sensing. International Journal of Remote Sensing, 18(4), 699-709.

Blackard, J.A., Dean, D.J. 1999. Comparative accuracies of artificial neural networks and discriminant analysis in predicting forest cover types from cartographic variables. Computers and Electronics in Agriculture, 24(3), 131-151.

Cohen, J. 1960. A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20(1), 37-46.

Dalponte, M. Bruzzone, L., Gianelle, D. 2008. Fusion of hyperspectral and LIDAR remote sensing data for classification of complex forest areas. IEEE Transactions in Geosciences and Remote Sensing, 46(5), 1416-1427.

Falkowski, M.J., Evans, J.S., Martinuzzi, S., Gessler, P.E., Hudak, A.T. 2009. Characterizing forest succession with lidar data: An evaluation for the Inland Northwest, USA. Remote Sensing of Environment, 113(5), 946-956.

Falkowski, M.J., Hudak, A.T., Crookston, N.L., Gessler, P.E., Uebler, E.H., Smith, A.M.S. 2010. Landscapescale parameterization of a tree-level forest growth model: A k-nearest neighbor imputation approach incorporating LiDAR data. Canadian Journal of Forest Research, 40(2), 184-199.

Foody, G.M., Campbell, N.A., Trodd, N.M., Wood, T.F. 1992. Derivation and applications of probabilistic measures of class membership from the maximumlikelihood classification. Photogrammetric Engineering & Remote Sensing, 58(9), 1335-1341.

García, M., Riaño, D., Chuvieco, E., Salas, J., Danson, F.M. 2011. Multispectral and LIDAR Data Fusion for Fuel Type Mapping Using Support Vector Machine and Decision Rules. Remote Sensing of

Environment, 115(6), 1369-1379.

García-Gutiérrez, J., Martínez-Álvarez, F., Riquelme, J.C. 2010. Using Remote Data Mining on LIDAR and Imagery Fusion Data to Develop Land Cover Maps. Trends in Applied Intelligent Systems. Lecture Notes in Computer Science, 6096, 378-387.

Geerling, G.W., Labrador-Garcia, M., Clevers, J.G.P.W., Ragas, A.M.J., Smits, A.J.M. 2007. Classification of floodplain vegetation by data fusion of spectral (CASI) and LiDAR data. International Journal of Remote Sensing, 28(19), 4263-4284.

Hastie, T., Tibshirani, R., Friedman, J.H. 2009. The Elements of Statistical Learning. Data Mining, Inference, and Prediction. New York: Springer

Heikkinen, V., Korpela, I., Tokola, T., Honkavaara, E. Parkkinen, J. 2011. An SVM Classification of Tree Species Radiometric Signatures Based on the Leica ADS40 Sensor. IEEE Transactions on Geoscience and Remote Sensing, 49(11), 4539-4551.

Holmgren, J., Persson, Å., Södermann, U. 2008. Species identification of individual trees by combining high resolution LiDAR data with multi-spectral images. International Journal of Remote Sensing, 29(5), 1537-1552.

Hudak, A.T., Crookston, N.L., Evans, J.S., Hall, D.E., Falkowski, M.J. 2008. Nearest neighbor imputation of species-level, plot-scale forest structure attributes from LiDAR data. Remote Sensing of Environment, 112(5), 2232-2245.

Kilkki, P., Päivinen, R. 1987. Reference sample plots to combine field measurements and satellite data in forest inventories. University of Helsinki, Department of Forest Mensuration and Management, Research Notes, 19, 209-215.

Kim, S., McGaughey, R.J., Andersen, H.E., Schreuder, G. 2009. Tree Species Differentiation Using Intensity Data Derived From Leaf-on and Leaf-Off Airborne Laser Scanner Data. Remote Sensing of Environment, 113(8), 1575-1586.

Korpela, I., Ørka, H.O., Maltamo, M., Tokola, T., Hyyppä, J. 2010. Tree species classification in airborne LiDAR data: influence of stand and tree factors, intensity normalization and sensor type. Silva Fennica, 44(2), 319-339.

Liaw, A., Wiener, M. 2002. Classification and Regression by Random Forest. R News, 2(3), 18-22.

Maltamo, M., Malinen, J., Packalén, P., Suvanto, A., Kangas, J. 2006. Non-parametric estimation of stem volume using laser scanning, aerial photography and stand register data. Canadian Journal of Forest Research, 36(2), 426-436.

Maltamo, M., Malinen, J., Packalén, P., Suvanto, A., Kangas, J. 2006. Non-parametric estimation of stem volume using laser scanning, aerial photography and stand register data. Canadian Journal of Forest Research, 36(2), 426-436.

Maltamo, M., Kallio, E., Bollandsås, O.M., Næsset, E., Gobakken, T., Pesonen, A. 2014 Assessing dead wood by airborne laser scanning. In: Maltamo M., Næsset E., Vauhkonen J. (Eds.) Forestry Applications of Airborne Laser Scanning. Concepts and Case Studies. Managing Forest Ecosystems, 27. Dordrecht: Springer. pp. 375-395.

Martinuzzi, S., Vierling, L.A. Gould, W.A. Falkowski, M.J. Evans, J.S. Hudak, A.T., Vierling, K.T. 2009. Mapping snags and understory shrubs for a LiDAR based assessment of wildlife habitat suitability.Remote Sensing of Environment, 113(12), 2533-2546.

Næsset, E. 2002. Predicting forest stand characteristics with airborne scanning laser using a practical twostage procedure and field data. Remote Sensing of Environment, 80(1), 88-99.

McGaughey, R.J. 2012. FUSION/LDV: Software for LIDAR Data Analysis and Visualization. Version 3.10. Pacific Northwest Research Station. Seattle, WA: USDA Forest Service.

McInerney, D.O., Suarez-Minguez, J., Valbuena, R., Nieuwenhuis, M. 2010. Forest canopy height retrieval using LiDAR data, medium-resolution satellite imagery and kNN estimation in Aberfoyle, Scotland. Forestry, 83(2), 195-206.

McRoberts, R.E. 2012. Estimating forest attribute parameters for small areas using nearest neighbors techniques. Forest Ecology and Management, 272, 3-12.

Meyer, D., Dimitriadou, E., Hornik, K., Weingessel, A., Leisch, F. 2014a. e1071: Misc Functions of the Department of Statistics, TU Wien. R package version 1.6-4.

Meyer, D., Zeileis, A., Hornik, K. 2014b. vcd: Visualizing Categorical Data. R package version 1.3-2

Niska, H., Skön, J.P., Packalén, P., Tokola, T., Maltamo, M., Kolehmainen, M. 2010. Neural networks for the prediction of species specific plot volumes using airborne laser scanning and aerial photographs. IEEE Transactions on Geoscience and Remote Sensing, 48(3), 1076-1085.

NLS - National Land Survey of Finland 2013. Laserscanning data. Available at (accessed on September, 2013).

Ørka, H.O., Næsset, E., Bollandsås, O.M. 2009. Classifying Species of Individual Trees by Intensity and Structure Features Derived From Airborne Laser Scanner Data. Remote Sensing of Environment, 113(6), 1163-1174.

Ørka, H.O., Gobakken, T., Næsset, E., Ene, L., Lien, V. 2012. Simultaneously acquired airborne laser scanning and multispectral imagery for individual tree species identification. Canadian Journal of Remote Sensing, 38(2), 125-138.

Reese, H., Nyström, M., Nordkvist, K., Olsson, H. 2014. Combining airborne laser scanning data and optical satellite data for classification of alpine vegetation. International Journal of Applied Earth Observation and Geoinformation, 27, 81-90.

Ripley, B.D. 1996. Pattern Recognition and Neural Networks. Cambridge, UK: Cambridge University Press.

Rogan, J., Franklin, J., Stow, D., Miller, J., Woodcock, C., Roberts, D. 2008. Mapping land-cover modifications over large areas: A comparison of machine learning algorithms. Remote Sensing of Environment, 112(5), 2272-2283.

Rousseeuw, P.J. 1986. Multivariate estimation with high breakdown point. In: Grossman, W., Pflug, G., Vincze, I., Wertz, W. (eds.) Mathematical Statistics and Applications, 283-297. Dordrecht: Reidel.

Tomppo, E. 1993. National Forest Inventory of Finland. In: Nyyssönen, A. (ed.). Proceedings of FAO/ECE meeting of experts on global forest resources assessment, Kotka, Finland. The Finnish Forest Research Institute, Research Papers, 469, 204-205.

Torresan, C., Strunk, J., Zald, H.S.J, Zhiqiang, Y., Cohen, W.B. 2014. Comparing statistical techniques to classify the structure of mountain forest stands using CHM-derived metrics in Trento province (Italy). European Journal of Remote Sensing, 47, 75-94.

Valbuena, R. 2014. Integrating Airborne Laser Scanning with Data from Global Navigation Satellite Systems and Optical Sensors. In: Maltamo M., Næsset E. and Vauhkonen J. (Eds.) Forestry Applications of Airborne Laser Scanning. Concepts and Case Studies. Managing Forest Ecosystems, 27. Dordrecht: Springer. pp. 63-88.

Valbuena, R., Vauhkonen, J., Packalen, P., Pitkänen, J., Maltamo, M. 2014. Comparison of airborne laser scanning methods for estimating forest structure indicators based on Lorenz curves. ISPRS Journal of Photogrammetry and Remote Sensing, 95, 23-33.

Valbuena, R., Mehtätalo, L., Maltamo, M., Packalen, P. 2015. A rule-based method using L-moments for direct detection of shade-intolerant regeneration in Boreal forests from airborne laser scanning. Silvilaser 2015. La Grande Motte. France. pp. 47-49. Available at: (accessed on January, 2016).

Valbuena, R., Maltamo, M., Packalen, P. 2016. Classification of multi-layered forest development classes from low-density national airborne lidar datasets. Forestry, 89.

Van Aardt, J.A.N., Wynne, R.H., Scrivani, J.A. 2008. Lidar-based mapping of forest volume and biomass by taxonomic group using structurally homogenous segments. Photogrammetric Engineering and Remote Sensing, 74(8), 1033-1044.

Vauhkonen, J., Korpela, I., Maltamo, M., Tokola, T. 2010. Imputation of single-tree attributes using airborne laser scanning-based height, intensity, and alpha shape metrics. Remote Sensing of Environment, 114(6), 1263-1276.

Venables, W.N., Ripley, B.D. 2002. Modern Applied Statistics with S-Plus. New York: Springer.

Warnes, G.R. 2013. gmodels: Various R programming tools for model fitting. R package version

Yu, X., Hyyppä, J., Vastaranta, M., Holopainen, M., Viitala, R. 2011. Predicting individual tree attributes from airborne laser point clouds based on the random forests technique. ISPRS Journal of Photogrammetry and Remote Sensing, 66(1), 28-37.

Zhang, Z., Liu, X., Peterson, J., Wendy, W. 2011. Cool temperate rainforest and adjacent forests classification using airborne LiDAR data. Area, 43(4), 438-448.