Physically explainable CNN for SAR image classification

• Published in: ISPRS journal of photogrammetry and remote sensing Vol. 190; pp. 25 - 37
• Main Authors: Huang, Zhongling, Yao, Xiwen, Liu, Ying, Dumitru, Corneliu Octavian, Datcu, Mihai, Han, Junwei
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.08.2022
• Subjects: data collection; Explainable deep learning; hybrids; image interpretation; neural networks; photogrammetry; Physical model; physics; prediction; Prior knowledge; SAR image classification; synthetic aperture radar; Prior knowledge; SAR image classification; Physical model; Explainable deep learning
• ISSN: 0924-2716; 1872-8235
• DOI: 10.1016/j.isprsjprs.2022.05.008

Integrating the special electromagnetic characteristics of Synthetic Aperture Radar (SAR) in deep neural networks is essential in order to enhance the explainability and physics awareness of deep learning. In this paper, we first propose a novel physically explainable convolutional neural network fo...