Two-stage subpixel impervious surface coverage estimation: comparing classification and regression trees and artificial neural networks

The paper presents accuracy comparison of subpixel classification based on medium resolution Landsat images, performed using machine learning algorithms built on decision and regression trees method (C.5.0/Cubist and Random Forest) and artificial neural networks. The aim of the study was to obtain the pattern of percentage impervious surface coverage, valid for the period of 2009-2010. Imperviousness index map generation was a two-stage procedure. The first step was classification, which divided the study area into categories: i) completely permeable (imperviousness index less than 1%) and ii) fully or partially impervious areas. For pixels classified as impervious, the percentage of impervious surface coverage in pixel area was estimated. The root mean square errors (RMS) of determination of the percentage of the impervious surfaces within a single pixel were 11.0% for C.5.0/Cubist method, 11.3% for Random Forest method and 12.6% using artificial neural networks. The introduction of the initial hard classification into completely permeable areas (with imperviousness index <1%) and impervious areas, allowed to improve the accuracy of imperviousness index estimation on poorly urbanized areas covering large areas of the Dobczyce Reservoir catchment. The effect is also visible on final imperviousness index maps.

[1]  Rozpoznanie form pokrycia i użytkowania ziemi na zdjęciu satelitarnym Landsat ETM+ metodą klasyfikacji obiektowej , 2006 .

[2]  Philip S. Yu,et al.  Top 10 algorithms in data mining , 2007, Knowledge and Information Systems.

[3]  Qihao Weng,et al.  Remote sensing of impervious surfaces in the urban areas: Requirements, methods, and trends , 2012 .

[4]  Paul M. Mather,et al.  The use of backpropagating artificial neural networks in land cover classification , 2003 .

[5]  Changshan Wu,et al.  The use of single-date MODIS imagery for estimating large-scale urban impervious surface fraction with spectral mixture analysis and machine learning techniques , 2013 .

[6]  Yang Shao,et al.  Comparison of support vector machine, neural network, and CART algorithms for the land-cover classification using limited training data points , 2012 .

[7]  W. Drzewiecki.  Sub-pixel classification of middle-resolution satellite images – evaluation of regression trees applicability to monitor impervious surfaces coverage , 2010 .

[8]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[9]  Scott L. Powell,et al.  Quantification of impervious surface in the Snohomish Water Resources Inventory Area of Western Washington from 1972–2006 , 2007 .

[10]  Jeffrey T. Walton.  Subpixel urban land cover estimation: comparing cubist, random forests, and support vector regression , 2008 .

[11]  Hanqiu Xu.  Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery , 2006 .

[12]  D. Civco,et al.  IMPERVIOUS SURFACE MAPPING FOR THE STATE OF CONNECTICUT 1 , 1997 .

[13]  Deepak Khare,et al.  Monitoring and modelling of urban sprawl using remote sensing and GIS techniques , 2008, Int. J. Appl. Earth Obs. Geoinformation.

[14]  C. Ji.  Land-use classification of remotely sensed data using Kohonen Self-Organizing Feature Map neural networks. , 2000 .

[15]  Max Kuhn,et al.  Applied Predictive Modeling , 2013 .

[16]  Sangbum Lee,et al.  Subpixel analysis of Landsat ETM/sup +/ using self-organizing map (SOM) neural networks for urban land cover characterization , 2006, IEEE Transactions on Geoscience and Remote Sensing.

[17]  Marvin E. Bauer,et al.  Multi‐level Land Cover Mapping of the Twin Cities (Minnesota) Metropolitan Area with Multi‐seasonal Landsat TM/ETM+ Data , 2005 .

[18]  G. Xian,et al.  Mapping impervious surfaces using classification and regression tree algorithm , 2007 .

[19]  Ning Qian,et al.  On the momentum term in gradient descent learning algorithms , 1999, Neural Networks.

[20]  F. Canters,et al.  Improving Distributed Runoff Prediction in Urbanized Catchments with Remote Sensing based Estimates of Impervious Surface Cover , 2008, Sensors.

[21]  Robert A. Jacobs,et al.  Increased rates of convergence through learning rate adaptation , 1987, Neural Networks.

[22]  C. Deng,et al.  BCI: A biophysical composition index for remote sensing of urban environments , 2012 .

[23]  R. Hecht-Nielsen,et al.  Theory of the Back Propagation Neural Network , 1989 .

[24]  Qihao Weng,et al.  Medium Spatial Resolution Satellite Imagery for Estimating and Mapping Urban Impervious Surfaces Using LSMA and ANN , 2008, IEEE Transactions on Geoscience and Remote Sensing.

[25]  D. Civco,et al.  SUBPIXEL IMPERVIOUS SURFACE MAPPING , 2001 .

[26]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[27]  Xuefei Hu,et al.  Estimating impervious surfaces from medium spatial resolution imagery using the self-organizing map and multi-layer perceptron neural networks. , 2009 .

[28]  J. Wickham,et al.  Completion of the 2001 National Land Cover Database for the conterminous United States , 2007 .

[29]  W. Drzewiecki,et al.  Application of Landsat imagery based vegetation indices to imperviousness index mapping , 2009 .