RSOD Dataset is a dataset for object detection in remote sensing images. It contains four types of targets: airplanes, playgrounds, overpasses, and oil drums. The number of images is: 446 images - 4993 airplanes, 189 images - 191 playgrounds, 176 images - 180 overpasses, and 165 images - 1586 oil drums.

This dataset was released by Wuhan University in 2015. Related papers include "Elliptic Fourier transformation-based histograms of oriented gradients for rotationally invariant object detection in remote-sensing images" and "Accurate Object Localization in Remote Sensing Images Based on Convolutional Neural Networks".