Favoritar este trabalho

This paper describes the process of multisensor time series composition intended for land-use and land-cover classification. Our objective is to verify the robustness of the Time Weighted Dynamic Time Warp algorithm when applied on images from multiples sensors. Our approach increased temporal resolution to improve the quality of classification. Our data was acquired between 2000 and 2016 and a small area of Mato Grosso, Brazil was analyzed. The results utilizing multisensor composition are promising and consistent with early results of lower resolution images.