Data Mining, A Promising Tool for Large-Area Cropland Mapping

The northern fringe of sub-Saharan Africa is a region that is considered to be particularly vulnerable to climate variability and change, and it is a location in which food security remains a major challenge. To address these issues, it is essential to develop global data sets of the geographic distribution of agricultural land use. The objectives of this study were to test an original data mining approach for classifying and mapping the cropped land in West Africa using coarse-resolution imagery and to compare the classification results with those obtained from a classic ISODATA approach. The data mining approach is able to handle large volumes of data and is based on different descriptors (65) of the land use, including the spatial and temporal satellite-derived metrics of 12 MODIS NDVI 16-day composite images and the static attributes taken from field surveys. The classic ISODATA method showed that 68.3% of pixels from a SPOT reference map were correctly classified in three validation sites versus 57.8% for the data mining approach. Validation by field observations showed equivalent results for both methods with an F-score of 0.72. The results of this study demonstrated the relevance of the use of data-mining tools for large-area monitoring.

[1]  Mikhail F. Kanevski,et al.  A Survey of Active Learning Algorithms for Supervised Remote Sensing Image Classification , 2011, IEEE Journal of Selected Topics in Signal Processing.

[2]  Clement Atzberger,et al.  Evaluation of Sentinel-2 Spectral Sampling for Radiative Transfer Model Based LAI Estimation of Wheat, Sugar Beet, and Maize , 2011, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[3]  Jianyong Wang,et al.  Mining sequential patterns by pattern-growth: the PrefixSpan approach , 2004, IEEE Transactions on Knowledge and Data Engineering.

[4]  Martin Herold,et al.  Some challenges in global land cover mapping : An assessment of agreement and accuracy in existing 1 km datasets , 2008 .

[5]  A. J. Stern,et al.  Application of MODIS derived parameters for regional crop yield assessment , 2005 .

[6]  Danny Lo Seen,et al.  Crop area mapping in West Africa using landscape stratification of MODIS time series and comparison with existing global land products , 2012, Int. J. Appl. Earth Obs. Geoinformation.

[7]  D. D. Krezhova,et al.  Development and testing of a statistical texture model for land cover classification of the Black Sea region with MODIS imagery , 2010 .

[8]  Christophe Rigotti,et al.  Unsupervised Spatiotemporal Mining of Satellite Image Time Series Using Grouped Frequent Sequential Patterns , 2011, IEEE Transactions on Geoscience and Remote Sensing.

[9]  Gilberto Câmara,et al.  Mining patterns of change in remote sensing image databases , 2005, Fifth IEEE International Conference on Data Mining (ICDM'05).

[10]  Clement Atzberger,et al.  Object Based Image Analysis and Data Mining applied to a remotely sensed Landsat time-series to map sugarcane over large areas , 2012 .

[11]  Prasad S. Thenkabail,et al.  Ganges and Indus river basin land use/land cover (LULC) and irrigated area mapping using continuous streams of MODIS data , 2005 .

[12]  Maguelonne Teisseire,et al.  Recherche de motifs et cartographie des surfaces agricoles. Des relevés de terrain aux données satellitaires : application au Mali , 2011, Rev. Int. Géomatique.

[13]  Michael A. Wulder,et al.  Landsat continuity: Issues and opportunities for land cover monitoring , 2008 .

[14]  William A. Salas,et al.  Monitoring Rice Agriculture in the Sacramento Valley, USA With Multitemporal PALSAR and MODIS Imagery , 2011, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[15]  Pierre Gançarski,et al.  A global averaging method for dynamic time warping, with applications to clustering , 2011, Pattern Recognit..

[16]  J. Mustard,et al.  Wavelet analysis of MODIS time series to detect expansion and intensification of row-crop agriculture in Brazil , 2008 .

[17]  C. Burnett,et al.  A multi-scale segmentation/object relationship modelling methodology for landscape analysis , 2003 .

[18]  Maguelonne Teisseire,et al.  Mining Sequential Patterns from MODIS Time Series for Cultivated Area Mapping , 2011, AGILE Conf..

[19]  S. Goward,et al.  Global Primary Production: A Remote Sensing Approach , 1995 .

[20]  Steffen Fritz,et al.  Geo-Wiki.Org: The Use of Crowdsourcing to Improve Global Land Cover , 2009, Remote. Sens..

[21]  D. Lobell,et al.  Cropland distributions from temporal unmixing of MODIS data , 2004 .

[22]  Wei Ding,et al.  Discovery of Geospatial Discriminating Patterns from Remote Sensing Datasets , 2009, SDM.

[23]  Philippe Bolon,et al.  On Extracting Evolutions from Satellite Image Time Series , 2008, IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium.

[24]  B. Wardlow,et al.  Large-area crop mapping using time-series MODIS 250 m NDVI data: An assessment for the U.S. Central Great Plains , 2008 .

[25]  Germain Forestier,et al.  Analysing Satellite Image Time Series by Means of Pattern Mining , 2010, IDEAL.

[26]  Terry L. Kastens,et al.  Image masking for crop yield forecasting using AVHRR NDVI time series imagery , 2005 .

[27]  M. Keller,et al.  Remote sensing of selective logging in Amazonia: Assessing limitations based on detailed field observations, Landsat ETM+, and textural analysis , 2002 .

[28]  B. Wardlow,et al.  Analysis of time-series MODIS 250 m vegetation index data for crop classification in the U.S. Central Great Plains , 2007 .

[29]  Chris Funk,et al.  Crop area estimation using high and medium resolution satellite imagery in areas with complex topography , 2008 .

[30]  P. Defourny,et al.  Retrieving forest structure variables based on image texture analysis and IKONOS-2 imagery , 2006 .

[31]  P. C. Doraiswamya,et al.  Crop condition and yield simulations using Landsat and MODIS , 2004 .

[32]  Florent Masseglia,et al.  The PSP Approach for Mining Sequential Patterns , 1998, PKDD.

[33]  J. Wickham,et al.  Effects of landscape characteristics on land-cover class accuracy , 2003 .

[34]  Anne Laurent,et al.  M2SP: Mining Sequential Patterns Among Several Dimensions , 2005, PKDD.

[35]  Zoran Obradovic,et al.  Modeling Multivariate Spatio-Temporal Remote Sensing Data with Large Gaps , 2011, IJCAI.

[36]  Carla E. Brodley,et al.  Redefining class definitions using constraint-based clustering: an application to remote sensing of the earth's surface , 2010, KDD.

[37]  Dino Ienco,et al.  LODE: A distance-based classifier built on ensembles of positive and negative observations , 2012, Pattern Recognit..

[38]  I. Herrmann,et al.  LAI assessment of wheat and potato crops by VENμS and Sentinel-2 bands , 2011 .

[39]  R. Suganya,et al.  Data Mining Concepts and Techniques , 2010 .

[40]  Christophe Rigotti,et al.  Efficient Spatio-temporal Mining of Satellite Image Time Series for Agricultural Monitoring , 2012, Trans. Mach. Learn. Data Min..

[41]  Damien Arvor,et al.  Analyzing the agricultural transition in Mato Grosso, Brazil, using satellite-derived indices , 2012 .

[42]  Rajeev Motwani,et al.  Dynamic itemset counting and implication rules for market basket data , 1997, SIGMOD '97.