Multisource Remotely Sensed Wetland Information Exploration Using Interactive Visualization Methods

Abstract

In wetland research, the increasing amount of data available from earth observation technologies provides new challenges for data analysis and wetland mapping. Interactive visualization can provide new techniques for knowledge discovery in remote sensing data. This paper analyzes the characteristics of remotely sensed wetland data and introduces a group of spatial-attribute interactive multidimensional visualization methods to browse the parameter space of remotely sensed wetland information. A case study using multisource remote sensing data for a wetland in the Daxinganling region shows that bands selection through visualization methods can help with the exploration of the positional and spectral characteristics of sample pixels.