An Object-Based Classification Approach for Mapping Migrant Housing in the Mega-Urban Area of the Pearl River Delta (China)

Remote Sensing

Sebastian d’Oleire-Oltmanns¹  Bodo Coenradie¹ 

[1] Geoinformation in Environmental Planning, Technische Universität Berlin, D-10623 Berlin, Germany; E-Mails:

关键词: urban structure types;    object-based classification;    land-use change;    SPOT5;    urban sprawl;   

DOI  :  10.3390/rs3081710

来源: mdpi

【 摘 要 】

Urban areas develop on formal and informal levels. Informal development is often highly dynamic, leading to a lag of spatial information about urban structure types. In this work, an object-based remote sensing approach will be presented to map the migrant housing urban structure type in the Pearl River Delta, China. SPOT5 data were utilized for the classification (auxiliary data, particularly up-to-date cadastral data, were not available). A hierarchically structured classification process was used to create (spectral) independence from single satellite scenes and to arrive at a transferrable classification process. Using the presented classification approach, an overall classification accuracy of migrant housing of 68.0% is attained.

【 授权许可】

CC BY   
© 2011 by the authors; licensee MDPI, Basel, Switzerland.