The open-pit mines extraction and change detection dataset (OMECSet) and the code of CTMNet

This dataset is a self-made open-pit mines extraction and change detection dataset with image data from the 2019, 2020, and 2021 Gaofen-6 VHR images, and mask files are produced by manual visual interpretation. Through cropping, the training dataset contains 660 sub-image pairs, each consisting of a pre-temporal remote sensing image file and the corresponding open-pit mine mask file, post-temporal remote sensing image file and the corresponding open-pit mine mask file, and a change area mask file. It reaches 2640 sub-image pairs after data augmentation. The testing dataset contains 212 sub-image pairs. All image pairs in the training dataset and test dataset are 256×256, with a resolution of 2 meters, and include 3 bands of R, G, B. If you find this work valuable or use our dataset in your research, please consider citing the paper ‘CTMNet: Enhanced Open-pit Mine Extraction and Change Detection with a Hybrid CNN-Transformer Multitask Network'.