Title: Automatic updating of an object-based tropical forest cover classification and change assessment
Authors: RASI RastislavBEUCHLE Rene'BODART CatherineVOLLMAR MICHAELSELIGER RomanACHARD Frederic
Citation: IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING vol. 6 no. 1 p. 66-73
Publisher: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
Publication Year: 2013
JRC N°: JRC72601
ISSN: 1939-1404
URI: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6332549
http://publications.jrc.ec.europa.eu/repository/handle/JRC72601
DOI: 10.1109/JSTARS.2012.2217733
Type: Articles in periodicals and books
Abstract: The TREES-3 project of the European Commission’s Joint Research Centre is producing estimates of tropical forest cover changes for two time periods: 1990–2000 and 2000–2010. This paper presents the method developed for the automatic change detection and classification of year 2010 imagery from the existing segmentation and classification results of first period 1990–2000. The imagery of year 2010 is processed in three steps: automatic segmentation, change detection and object spectral classification. The validated maps of forest cover changes for the period 1990–2000 are integrated in vector format as thematic input layer into year 2010 imagery segmentation and classification process. Object-based change detection technique is applied using Tasselled Cap components and spectral Euclidian Distances. Objects detected as changed are classified in two steps: supervised classification and change vector analysis for unclassified remaining objects. The training areas for supervised classification are selected as objects identified as ‘unchanged’ and spectral signatures are extracted from 2010 imagery. Change vectors are defined according to available classification of year 2000, by spectral differences of land cover classes from year 2000 imagery. The segmentation approach was tested on 568 sample units spread over Brazil. The segmentation results for year 2010 demonstrated consistency with segmentation of imagery for the period 1990-2000. The resulting overall accuracy of this automatic classification was estimated for the 281 sample units of Brazilian Amazonian region and for 201 sample units of three more complex biomes at 92% and 91% respectively.
JRC Directorate:Sustainable Resources

Files in This Item:
There are no files associated with this item.


Items in repository are protected by copyright, with all rights reserved, unless otherwise indicated.