Optimized Minimum-Search for SAR Backprojection Autofocus on GPUs Using CUDA

Autofocus techniques for synthetic aperture radar (SAR) can improve the image quality substantially. Their high computational complexity imposes a challenge when employing them in runtime-critical implementations. This paper presents an autofocus implementation for stripmap SAR specially optimized for parallel architectures like GPUs. Thorough evaluation using real SAR data shows that the tunable parameters of the algorithm allow to counterbalance runtime and achieved image quality.© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works