Michael Schmitt, Claudio Persello, Gemine Vivone, Dalton Lunga, Wenzhi Liao, Naoto Yokoya, Pedram Ghamisi, and Ronny Hänsch
The Image Analysis and Data Fusion Technical Committee (IADF TC) of the IEEE Geoscience and Remote Sensing Society (GRSS) serves as a global, multidisciplinary network for geospatial image analysis (IA), eg, machine learning (ML), deep learning (DL), image and signal processing, and big data and data fusion (DF), eg, multisensor, multiscale, and multitemporal data integration.
The wide scope of IADF broadened further with the recent advances of IA and DF. The increasing availability of space or airborne remote sensing data requires advanced methods for multimodal DF, including image alignment, low-level preprocessing, feature extraction, and image fusion. It also allows for new opportunities for IA such as sensor transcoding, cross-modal image interpretation, and domain adaptation. Many of these advancements are achieved by ML approaches. In particular, DL has been extremely successful at all levels of the processing chain to automatically interpret remote sensing images including image compression, cloud removal, noise suppression, image registration and normalization, feature extraction, and representation learning as well as semantic analysis and extraction of bio-/geophysical parameters. One of the main causes for the recent success of ML and DL is the increasing availability of large-scale benchmark data sets, i.e., data that contains reference data additionally to the input images, which allows the proper training and evaluation of learning-based approaches.
To ensure that these diverse aspects are well represented by IADF, three working groups (WGs) were created that are dedicated to distinct fields within the scope of IA and DF, namely benchmarking (WG-BEN), image and signal processing (WG-ISP), and ML/DL for IA (WG-MIA).
IEEE Geoscience and Remote Sensing Magazine, 9, 3, 165-166, 2021-09-29.