Using Landsat and nighttime lights for supervised pixel-based image classification of urban land cover

REMOTE SENSING OF ENVIRONMENT

卷:205

Article

Goldblatt, Ran¹  Stuhlmacher, Michelle F.²  Tellman, Beth²  Clinton, Nicholas³  Hanson, Gordon¹  Georgescu, Matei²  Wang, Chuyuan²  Serrano-Candela, Fidel⁴  Khandelwal, Amit K.⁵  Cheng, Wan-Hwa²  Balling, Robert C., Jr.² 

[1] Univ Calif San Diego, Sch Global Policy & Strategy, 9500 Gilman Dr, La Jolla, CA 92093 USA

[2] Arizona State Univ, Sch Geog Sci & Urban Planning, 976 S Forest Mall, Tempe, AZ 85281 USA

[3] Google Inc, 1600 Amphitheatre Pkwy, Mountain View, CA 94043 USA

[4] Univ Nacl Autonoma Mexico, Lab Nacl Ciencias Sostenibilidad, Apartado Postal 70-275 Ciudad Univ, Mexico City, DF, Mexico

[5] Columbia Univ, Columbia Business Sch, New York, NY 10027 USA

关键词: Urbanization;    Built-up land cover;    Nighttime light;    Image classification;    Google Earth Engine;   

DOI  :  10.1016/j.rse.2017.11.026

来源: Elsevier

【 摘 要 】

Reliable representations of global urban extent remain limited, hindering scientific progress across a range of disciplines that study functionality of sustainable cities. We present an efficient and low-cost machine-learning approach for pixel-based image classification of built-up areas at a large geographic scale using Landsat data. Our methodology combines nighttime-lights data and Landsat 8 and overcomes the lack of extensive ground reference data. We demonstrate the effectiveness of our methodology, which is implemented in Google Earth Engine, through the development of accurate 30 m resolution maps that characterize built-up land cover in three geographically diverse countries: India, Mexico, and the US. Our approach highlights the usefulness of data fusion techniques for studying the built environment and is a first step towards the creation of an accurate global-scale map of urban land cover over time.

【 授权许可】

Free