Wetland extraction of Yancheng coastal area based on ALOS data

Abstract

Jiangsu coastal area is abundant in wetland resources and possesses wetland reserve of global relevancy. However, the wetland area has been shrinking with more and more human activities. In this paper, 5 ALOS AVNIR-2 images as well as an additional TM image acquired at low tide level were selected to carry out a research of Yancheng coastal wetland. Characteristics of subcategories and vegetation coverage were surveyed during filed work before a 2-tier decision tree method was adopted to distinguish different wetland categories. At last, the total area was classified as 8 major wetland categories with an overall accuracy of 97.16% as well as Kappa coefficient of 0.96. Only 2 sampling sites biased in the classification when 28 field recorded samples were compared and the result was satisfying.