A Progressive Crop-Type Classification Using Multitemporal Remote Sensing Data and Phenological Information

Local parameters for climate modelling are highly dependent on crop types and their phenological growth stage. The land cover change of agricultural areas during the growing season provides important information to distinguish crop types. The presented progressive classification algorithm identifies crop types based on their phenological development and their corresponding reflectance characteristics in multitemporal satellite images of the four sensors Landsat-7 and -8, Sentinel-2A and RapidEye. It distinguishes crop types not only retrospectively, but progressively during the growing season starting in early spring. Binary fuzzy c-means clustering differentiates seven crop types in eight decisions at particular time periods. These decisions are based on expert knowledge about plant characteristics in different phenological stages. The unsupervised classification approach and previously defined decisions enable the algorithm to work independently of training data. The fuzzy approach provides certainties of crop-type existence and generates first results in early spring. The accuracy and reliability of the classification results improve with increasing time. The method is developed at the German Durable Environmental Multidisciplinary Monitoring Information Network (DEMMIN). The study area is an official test site of the Joint Experiment of Crop Assessment and Monitoring (JECAM) and is located in an intensely agricultural used area in Northern Germany. Classification results were produced for the growing seasons 2015 and 2016. The overall accuracy in 2015 amounted to 89.49%. A challenge remains the distinct separation of wheat and rye, whereas barley, rapeseed, potato, corn and sugar beet are classified with high accuracies. The overall accuracy in 2016 was lower (77.19%) due to unfavourable weather conditions.ZusammenfassungEine progressive Klassifikation von Fruchtarten unter Verwendung von multitemporalen Fernerkundungsdaten und phänologischen Informationen. Lokale Parameter für die Klimamodellierung sind stark von Fruchtarten und deren aktuellen phänologischen Entwicklungsstadien abhängig. Die Bodenbedeckung landwirtschaftlicher Flächen verändert sich im Laufe der Vegetationsperiode erheblich und liefert somit wichtige Informationen zur Unterscheidung verschiedener Fruchtarten. Der vorgestellte fortschreitende Klassifikationsalgorithmus identifiziert Fruchtarten anhand ihrer phänologischen Entwicklung und ihrer entsprechenden charakteristischen Reflexionseigenschaften von multitemporalen optischen Satellitenbildern der Sensoren Landsat-7 und -8, Sentinel-2 und RapidEye. Die Fruchtarten werden nicht am Ende der Vegetationsperiode klassifiziert, sondern fortschreitend mit Beginn der Vegetationsperiode im Frühling. Mithilfe eines binären Fuzzy C-Means Clustering können sieben Fruchtarten in acht Entscheidungsschritten zu bestimmten Zeiträumen unterschieden werden. Diese Entscheidungen basieren auf Expertenwissen über Pflanzencharakteristiken in den verschiedenen phänologischen Phasen. Die unüberwachte Klassifikationsmethode und die zuvor definierten Entscheidungen ermöglichen es dem Algorithmus, unabhängig von Trainingsdaten zu arbeiten. Die Verwendung von ”fuzzy” Entscheidungen erlaubt es, die Verlässlichkeitder Fruchtartenzuordnung anzugeben und liefert erste Ergebnisse bereits im Frühjahr. Die Genauigkeit und Zuverlässigkeit der Klassifikationsergebnisse verbessern sich im Laufe der Zeit. Die Methode wurde am Kalibrierungs- und Validierungsstandort DEMMIN (Durable Environmental Multidisciplinary Monitoring Information Network) entwickelt. Das Untersuchungsgebiet liegt in einem intensiv landwirtschaftlich genutzten Gebiet in Norddeutschland und ist außerdem offizieller Teststandort des Joint Experiment of Crop Assessment and Monitoring (JECAM). Klassifikationsergebnisse wurden für die Vegetationsperioden der Jahre 2015 und 2016 erstellt. Die Gesamtgenauigkeit im Jahr 2015 betrug 89,49%. Eine Herausforderung bleibt die eindeutige Trennung von Weizen und Roggen, wohingegen Gerste, Raps, Kartoffeln, Mais und Zuckerrüben mit hohen Genauigkeiten identifiziert werden konnten. Aufgrund ungünstiger Witterungsbedingungen war die Gesamtgenauigkeit für das Jahr 2016 mit 77,19% etwas geringer.

[1]  V. Simonneaux,et al.  The use of high‐resolution image time series for crop classification and evapotranspiration estimate over an irrigated area in central Morocco , 2008 .

[2]  Lucy Marshall,et al.  Object-oriented crop classification using multitemporal ETM+ SLC-off imagery and random forest , 2013 .

[3]  H. Bleiholder,et al.  The BBCH system to coding the phenological growth stages of plants - history and publications. , 2009 .

[4]  Clement Atzberger,et al.  First Experience with Sentinel-2 Data for Crop and Tree Species Classifications in Central Europe , 2016, Remote. Sens..

[5]  Clement Atzberger,et al.  Using Low Resolution Satellite Imagery for Yield Prediction and Yield Anomaly Detection , 2013, Remote. Sens..

[6]  Giles M. Foody,et al.  Crop classification by support vector machine with intelligently selected training data for an operational application , 2008 .

[7]  N. Ishitsuka,et al.  Crop discrimination with multitemporal SPOT/HRV data in the Saga Plains, Japan , 2001 .

[8]  M. Zortea,et al.  OBJECT-BASED CLOUD AND CLOUD SHADOW DETECTION IN LANDSAT IMAGES FOR TROPICAL FOREST MONITORING , 2012 .

[9]  Oguz Gungor,et al.  Evaluation of random forest method for agricultural crop classification , 2012 .

[10]  M. Bossard,et al.  CORINE land cover technical guide - Addendum 2000 , 2000 .

[11]  Mustafa Turker,et al.  Sequential masking classification of multi‐temporal Landsat7 ETM+ images for field‐based crop mapping in Karacabey, Turkey , 2005 .

[12]  Robert LIN,et al.  NOTE ON FUZZY SETS , 2014 .

[13]  Christopher O. Justice,et al.  Cloud cover throughout the agricultural growing season: Impacts on passive optical earth observations , 2015 .

[14]  Irena Hajnsek,et al.  A Network of Terrestrial Environmental Observatories in Germany , 2011 .

[15]  Xavier Blaes,et al.  Efficiency of crop identification based on optical and SAR image time series , 2005 .

[16]  Christopher Conrad,et al.  Impact of feature selection on the accuracy and spatial uncertainty of per-field crop classification using Support Vector Machines , 2013 .

[17]  Steven E. Franklin,et al.  A comparison of pixel-based and object-based image analysis with selected machine learning algorithms for the classification of agricultural landscapes using SPOT-5 HRG imagery , 2012 .

[18]  J. Six,et al.  Object-based crop identification using multiple vegetation indices, textural features and crop phenology , 2011 .

[19]  S. Cosentino,et al.  Dry matter and qualitative characteristics of alfalfa as affected by harvest times and soil water content , 2011 .

[20]  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 .

[21]  Raffaele Casa,et al.  Assessing Crop Water Demand by Remote Sensing and GIS for the Pontina Plain, Central Italy , 2009 .

[22]  Chi-Farn Chen,et al.  A Phenology-Based Classification of Time-Series MODIS Data for Rice Crop Monitoring in Mekong Delta, Vietnam , 2013, Remote. Sens..

[23]  Luis Guanter,et al.  Ready-to-Use Methods for the Detection of Clouds, Cirrus, Snow, Shadow, Water and Clear Sky Pixels in Sentinel-2 MSI Images , 2016, Remote. Sens..

[24]  S. Carpenter,et al.  Global Consequences of Land Use , 2005, Science.

[25]  Giorgos Mallinis,et al.  A Hidden Markov Models Approach for Crop Classification: Linking Crop Phenology to Time Series of Multi-Sensor Remote Sensing Data , 2015, Remote. Sens..

[26]  M. Mkhabela,et al.  Early maize yield forecasting in the four agro-ecological regions of Swaziland using NDVI data derived from NOAA's-AVHRR , 2005 .

[27]  Xiaojin Zhu,et al.  Crop Type Classification by Simultaneous Use of Satellite Images of Different Resolutions , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[28]  Russell G. Congalton,et al.  Assessing the accuracy of remotely sensed data : principles and practices , 1998 .

[29]  Anil K. Jain.  Data clustering: 50 years beyond K-means , 2008, Pattern Recognit. Lett..

[30]  Malcolm Davidson,et al.  Crop Classification Using Short-Revisit Multitemporal SAR Data , 2011, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[31]  Christopher Conrad,et al.  Satellite based calculation of spatially distributed crop water requirements for cotton and wheat cultivation in Fergana Valley, Uzbekistan , 2013 .

[32]  Jordi Inglada,et al.  Assessment of a Markov logic model of crop rotations for early crop mapping , 2015, Comput. Electron. Agric..

[33]  Claire Marais-Sicre,et al.  Improved Early Crop Type Identification By Joint Use of High Temporal Resolution SAR And Optical Image Time Series , 2016, Remote. Sens..

[34]  Tim R. McVicar,et al.  Determining temporal windows for crop discrimination with remote sensing: a case study in south-eastern Australia , 2004 .

[35]  Christopher Conrad,et al.  Integration of Optical and Synthetic Aperture Radar Imagery for Improving Crop Mapping in Northwestern Benin, West Africa , 2014, Remote. Sens..

[36]  Julie B. Odenweller,et al.  Crop identification using Landsat temporal-spectral profiles , 1984 .

[37]  Peter H Verburg,et al.  A review of global potentially available cropland estimates and their consequences for model‐based assessments , 2015, Global change biology.

[38]  Georg Bareth,et al.  Multi-Data Approach for remote sensing-based regional crop rotation mapping: A case study for the Rur catchment, Germany , 2017, Int. J. Appl. Earth Obs. Geoinformation.

[39]  Zhe Zhu,et al.  Object-based cloud and cloud shadow detection in Landsat imagery , 2012 .

[40]  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 .

[41]  G. Badhwar,et al.  Automatic corn-soybean classification using Landsat MSS data. I - Near-harvest crop proportion estimation. II - Early season crop proportion estimation , 1984 .

[42]  D. Bargiel,et al.  A new method for crop classification combining time series of radar images and crop phenology information. , 2017 .

[43]  H. Vereecken,et al.  Towards a network of observatories in terrestrial environmental research , 2006 .

[44]  Elizabeth Bradley.  Analysis of time series , 2003 .

[45]  Heather McNairn,et al.  TerraSAR-X and RADARSAT-2 for crop classification and acreage estimation , 2009, 2009 IEEE International Geoscience and Remote Sensing Symposium.

[46]  Saskia Foerster,et al.  Crop type mapping using spectral-temporal profiles and phenological information , 2012 .

[47]  Jean-Marie Monget,et al.  Corine Land Cover , 2018 .

[48]  Patrick Hostert,et al.  AROSICS: An Automated and Robust Open-Source Image Co-Registration Software for Multi-Sensor Satellite Data , 2017, Remote. Sens..

[49]  Francisco Zambrano Bigiarini.  Satellite Data for Agricultural Drought Analysis in Chile , 2017 .

[50]  A. J. W. De Wit,et al.  Efficiency and accuracy of per-field classification for operational crop mapping , 2004 .

[51]  Olena Dubovyk,et al.  Derivation of temporal windows for accurate crop discrimination in heterogeneous croplands of Uzbekistan using multitemporal RapidEye images , 2014 .

[52]  Giacomo Fontanelli,et al.  In-Season Mapping of Crop Type with Optical and X-Band SAR Data: A Classification Tree Approach Using Synoptic Seasonal Features , 2015, Remote. Sens..

[53]  J. Bezdek,et al.  FCM: The fuzzy c-means clustering algorithm , 1984 .

[54]  Tobias Landmann,et al.  Evaluating the sequential masking classification approach for improving crop discrimination in the Sudanian Savanna of West Africa , 2015, Comput. Electron. Agric..

[55]  Bing Zhang,et al.  Object-Based Crop Classification with Landsat-MODIS Enhanced Time-Series Data , 2015, Remote. Sens..