Predicting categorical forest variables using an improved k-Nearest Neighbour estimator and Landsat imagery

被引:60
作者
Tomppo, Erkki O. [1 ]
Gagliano, Caterina [2 ]
De Natale, Flora [2 ]
Katila, Matti [1 ]
McRoberts, Ronald E. [3 ]
机构
[1] Metla Finnish Forest Res Inst, FIN-00170 Helsinki, Finland
[2] Forest Monitoring & Planning Res Unit, Natl Council Agr Res, MPF, CRA, I-38100 Villazzano, Italy
[3] US Forest Serv, No Res Stn, St Paul, MN 55108 USA
关键词
National Forest Inventory; ik-NN; Categorical variables; Genetic algorithm; REMOTELY-SENSED DATA; ACCURACY; INVENTORY; ERROR; FIELD; ATTRIBUTES; AREA; BIAS;
D O I
10.1016/j.rse.2008.05.021
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
The k-Nearest Neighbour (k-NN) estimation and prediction technique is widely used to produce pixel-level predictions and areal estimates of continuous forest variables such as area and volume, often by sub-categories such as species. An advantage of k-NN is that the same parameters (e.g., k-value, distance metric, weight vector for the feature space variables) can be used for all variables, whether continuous or categorical. An obvious question is the degree to which accuracy can be improved if the k-NN estimation parameters are tailored for specific variable groups such as volumes by tree species or categorical variables. We investigated prediction of categorical forest attribute variables from satellite image spectral data using k-NN with optimisation of the weight vector for the ancillary variables obtained using a genetic algorithm. We tested several genetic algorithm fitness functions, all derived from well-known accuracy measures. For a Finnish test site, the categorical forest attribute variables were site fertility and tree species dominance, and for an Italian test site, the variables were forest type and conifer/broad-leaved dominance. The results for both test sites were validated using independent data sets. Our results indicate that use of the genetic algorithm to optimize the weight vector for prediction of a single forest attribute variable had a slight positive effect on the prediction accuracies for other variables. Errors can be further decreased if the optimisation is done by variable groups. (c) 2008 Elsevier Inc. All rights reserved.
引用
收藏
页码:500 / 517
页数:18
相关论文
共 68 条
  • [1] *AM SOC PHOT, 1983, MAN REM SENS
  • [2] [Anonymous], 2000, GLOBAL FOREST RESOUR
  • [3] [Anonymous], 2007, P 7 ANN FOR INV AN S
  • [4] [Anonymous], 1990, P SNS IUFRO WORKSH
  • [5] [Anonymous], 2003, VIENN DECL VIENN RES
  • [6] [Anonymous], 1997, Ph.D. Thesis
  • [7] [Anonymous], 2007, E43 COST
  • [8] Design-based approach to k-nearest neighbours technique for coupling field and remotely sensed data in forest surveys
    Baffetta, Federica
    Fattorini, Lorenzo
    Franceschi, Sara
    Corona, Piermaria
    [J]. REMOTE SENSING OF ENVIRONMENT, 2009, 113 (03) : 463 - 475
  • [9] BERTINI R, 2007, FORESTA, V4, P110
  • [10] Syndrome identification based on 2D analysis software
    Boehringer, Stefan
    Vollmar, Tobias
    Tasse, Christiane
    Wurtz, Rolf P.
    Gillessen-Kaesbach, Gabriele
    Horsthemke, Bernhard
    Wieczorek, Dagmar
    [J]. EUROPEAN JOURNAL OF HUMAN GENETICS, 2006, 14 (10) : 1082 - 1089