A SAR Dataset for ATR Development: the Synthetic and Measured Paired Labeled Experiment (SAMPLE)

摘要

The publicly-available Moving and Stationary Target Acquisition and Recognition (MSTAR) synthetic apertureradar (SAR) dataset has been an valuable tool in the development of SAR automatic target recognition(ATR) algorithms over the past two decades, leading to the achievement of excellent target classi cation results.However, because of the large number of possible sensor parameters, target con gurations and environmentalconditions, the SAR operating condition (OC) space is vast. This leads to the impossible task of collectingsu cient measured data to cover the entire OC space. Thus, synthetic data must be generated to augmentmeasured datasets. The study of synthetic data delity with respect to classi cation tasks is a non-trivial task.To that end, we introduce the Synthetic and Measured Paired and Labeled Experiment (SAMPLE) dataset,which consists of SAR imagery from the MSTAR dataset and well-matched synthetic data. By matching targetcon gurations and sensor parameters among the measured and synthetic data, the SAMPLE dataset is ideal forinvestigating the di erences between measured and synthetic SAR imagery. In addition to the dataset, we proposefour experimental designs challenging researchers to investigate the best ways to classify targets in measuredSAR imagery given synthetic SAR training imagery.