Change Detection of Mangroves at Subpixel Level of Synthesized Hyperspectral Data Using Multifractal Analysis Method

Abstract

Listen

Translate article

AbstractIn recent times rapid changes have been induced in mangrove forests due to anthropogenic activity and uncontrolled population growth. As a consequence distinct mangrove communities interact with each other. Interactions among the mangrove communities are competitive as well as mutual. In our study multifractal analysis method has been applied to assess the decadal changes in the mangrove zonation communities/classes, in this proposed method value of the degree of multifractality describes the in homogeneity or changes in the mangrove communities. In this paper, reconstructed high-resolution multispectral time-series data have been used to detect the changes in the communities. It has been observed that some mangrove communities are dominant over other mangrove communities, in our study area it has been noticed that the values of degree of multifractality are increasing for some mangrove communities such as Excoecaria agallocha dense and saline blank and are decreasing for Avicennia dense and Mixed Mangrove. The increased value of the degree of multifractality signifies the dominance of a particular mangrove community and the decreased value signifies less dominated mangrove community. The experimental results were endorsed through ground truth data. Thus, our proposed method can identify the dominance of different mangrove communities in the entire study area.KeywordsMultifractal analysisHyperspectralMultispectral