WHU-OHS Benchmark Dataset for Large-Scale Spectral Image Classification

The WHU-OHS dataset consists of 42 OHS satellite images acquired from more than 40 different locations in China. The images have a spatial resolution of 10 meters and a swath width of 60 kilometers. There are 32 spectral channels ranging from visible light to near infrared, with an average spectral resolution of 15 nanometers. We crop each image to 512 × 512 pixels with a step size of 32. There are 4822, 513, and 2,460 sub-images in the training set, validation set, and test set, respectively.

This dataset was released by Wuhan University in 2022, and the related paper results are "WHU-OHS : A benchmark dataset for large-scale Hersepctral Image classification".

Beijing Institute of Technology used this dataset to study and publish a paper in 2024.Dual-stage Hyperspectral Image Classification Model with Spectral Supertoken", has been accepted by ECCV 2024.