Dataset of simulated inverse synthetic aperture radar (ISAR) images of automotive targets.
This dataset consists of realistic simulated inverse synthetic aperture radar (ISAR) images of five commonly found automotive targets- a full-size car, a mid-size car, a bicycle, an auto-rickshaw, and a four-wheel truck. The ISAR images are generated of these targets moving along four distinct trajectories while performing four different maneuvers: right turn (South to East, East to North, North to West, and West to South); left turn (South to West, West to North, North to East, and East to South); U-turn (South to South, West to West, North to North, and East to East); and a straight trajectory (South to North; North to South, East to West, and West to East) at a four-way traffic junction. We have incorporated range-Doppler clutter and receiver noise in our simulated ISAR images.
This dataset is associated with the paper arXiv:2101.12535
Due to the large size of the dataset, we have divided it into five .zip files, corresponding to each target, namely: Autorikshwa.zip; Bicycle.zip; Fullsize_car.zip; Midsize_car.zip; and Truck.zip. Each *.zip file consists of five different signal to noise ratio (SNR) cases (+10, +5, 0, -5, -10 dB), and four range-Doppler based clutter cases (wind velocity +2.5, 5, 7.5 and 10 m/s). There are 16- different trajectories (four left turns, four right turns, four u-turns, and four straight trajectories) for each of these cases. We have approximately 45-49 images (.png and .mat format) corresponding to each case.