Synthetic Aperture Radar Target Recognition under Limited Training Data

This book reports the latest results in the study of synthetic aperture radar (SAR) automatic target recognition (ATR) and focuses on the theory and methods of limited-data SAR ATR, and experimental verification, and many other aspects. In practical applications, due to the scarcity of targets, the difficulty of SAR image acquisition, and the difficulty of accurate labeling. It is impossible to obtain sufficient SAR images. Thus, it is one of the core issues of concern in the field of SAR ATR. This book contains two main parts: classical limited-data SAR ATR and limited-data SAR causal ATR. The first part consists of theoretical foundations, and limited-data SAR ATR based on data augmentation and model design. The second part focuses on the theoretical foundations of SAR causal ATR, and methods based on causal feature extraction and causal intervention. This book also includes the research results of realization technology and experimental verification. Researchers, engineers, and graduate students in image processing and deep learning can benefit from this book, who want to learn the core theories, methods, and applications of SAR ATR.