Synthetic aperture radar images for cGAN training

Some novel methods for imaging based on synthetic aperture radar can result in images contaminated by artifacts as a consequence of pushing the limits of the algorithms. In order to mitigate the impact of this artifacts, image translation techniques can be exploited enabling to turn the SAR image into a cleaner one. For this purpose, multiple techniques can be used such as convolutional neural networks or generative adversial networks. However, the training of those systems can require a high number of images, which can be computationally expensive to generate. In this dataset, training images for two different problems are provided. The first one is related to This work presents an approach to enhance the quality of high-resolution images obtained by means of systems relying on synthetic aperture radar (SAR) . For this purpose, a deep learning method called conditional generative adversarial networks (cGAN) is applied to the imager outcome when it is prone to suffer artifacts. This is specially the case of novel systems pushing the limits of SAR (e.g., irregular sampling, multilayered media, etc.) resulting in very chaotic clutter and image artifacts that cannot be easily removed with conventional approaches. The cGAN can be trained to detect high-level  haracteristic features inthe image (e.g., parts of a scissor blade) so another output based on these detected features can be tailored. In other words, it can translate features contaminated by artifacts into clean features, effectively improving the quality of SAR images. Unlike other deep learning approaches, the training of the involved neural networks tends to be stable thanks to the structure based on two competing subsystems. The proposed approach is illustrated using simulated and measurement data in the context of two advanced near-field SAR systems considering: i) cylindricalmulti-layered media, and ii) freehand acquisitions. Results show that cGANs clearly outperforme conventional approaches removing most of the artifacts, enabling to produce a clean output image.