UC Merced Land-Use Dataset is a 21-class land use image remote sensing dataset for research. It has 100 categories of images, all extracted from the USGS National Map Urban Area Imagery series, which are used in urban areas across the country. The pixel resolution of the public domain images in this dataset is 1 foot, the image pixel size is 256*256, and it contains a total of 2100 scene images of 21 categories, with 100 images in each category.

This dataset was released by the UC Merced Computer Vision Laboratory in 2010, and the related paper is "Bag-Of-Visual-Words and Spatial Extensions for Land-Use Classification".