Can we use deep learning models to identify the functionality of plastics from space?

Shanyu Zhou, Lichao Mou, Yuansheng Hua, Lixian Zhang, Hermann Kaufmann, Xiao Xiang Zhu

Abstract: Highlights•First-time classification of the functions of plastics based on Sentinel-2 imagery.•Comparison of three basic deep learning models for classifying the functions of plastics.•Two-stage workflow for integrating hyperspectral PRISMA data with Sentinel-2 RGB images.•VGG-13 model reached an overall accuracy of 78%.