Mapping vegetation in an urban area with stratified classification and multiple endmember spectral mixture analysis

Vegetation is a fundamental element in urban ecosystems, and vegetation mapping is critical for urban and landscape planning and management. While remote sensing has increasingly been used for vegetation mapping, this spatially explicit approach can be challenging due to the spectral similarity between various vegetation types and the presence of complex features in the urban environment. The objective of this study is to develop a method that can help improve vegetation mapping in urban areas from medium-resolution remote sensor imagery. Central to our method is the combined use of stratified classification and multiple endmember spectral mixture analysis (MESMA) techniques. Specifically, we firstly partition the entire landscape into rural and urban subsets using road network density so that each subset can be processed independently to minimize the spectral confusion between some urban features and agricultural land covers. Secondly, we carefully extract all vegetation covers at the sub-pixel level for the urban subset by using the MESMA technique in order to account for small, fragmented vegetation patches that would be classified as non-vegetated classes otherwise. Thirdly, we adopt a separate supervised classification protocol to the rural subset and the vegetation covers extracted from the urban subset. Finally, we combine the classified outcomes from the two subsets to produce a complete map. We have implemented this method to produce a land cover map including various vegetation types from a Landsat Thematic Mapper (TM) image covering a large metropolitan area. It is found that this method has substantially outperformed two related ones that use the same supervised protocol to the entire area directly or to the rural subset and the urban subset without being MESMA processed. The advantage of our method is that it has extended the capability of sub-pixel analysis beyond vegetation abundance estimation and into the area of mapping thematic vegetation types in urban areas. (c) 2013 Elsevier Inc. All rights reserved.