Dataset Editor: Wang Zhirui  Sun Xian

High-resolution SAR Aircraft Detection and Recognition Dataset-1.0 (SAR-AIRcraft-1.0) releases 4368 images with a resolution of 1m. The imaging mode is the spotlight, and the polarization mode is the single polarization. The dataset contains images of four sizes, namely 800×800, 1000×1000, 1200×1200 and 1500×1500. It has 16,463 aircraft instances, including seven categories: A220, A320/321, A330, ARJ21, Boeing737, Boeing787 and other, which can support detection tasks, recognition tasks, and integrated detection and recognition tasks. This dataset has the characteristics of complex scenes, rich categories, dense targets, noise interference, diverse tasks, and multi-scale.

The format of the image in the dataset is JPEG, and the format of the annotation file is xml. The length and width of the image, the category of the target, and the location of the bounding boxes are provided in the corresponding annotation.

The data details can be found in the papers "SAR-AIRcraft-1.0: High-resolution SAR Aircraft Detection and Recognition Dataset" and "The instruction of SAR image object detection dataset.pdf".

The current data reference format is as follows:

WANG Zhirui, KANG Yuzhuo, ZENG Xuan, et al. SAR-AIRcraft-1.0: High-resolution SAR aircraft detection and recognition dataset[J]. Journal of Radars, 2023, 12(4): 906–922. doi:  10.12000/JR23043