Object Detection Datasets of Open-Pit Mining Areas in Northern and Southern China

Drawing upon existing land cover mapping efforts for open-pit mining areas distributed across distinct geographic zones, two representative study areas in northern and southern China were selected. Accordingly, two object detection datasets—the Northern Open-Pit Mining Area (NOMA) and the Southern Open-Pit Mining Area (SOMA)—were established. The NOMA dataset comprises 1,608 images with 2,180 labeled instances, whereas the SOMA dataset consists of 4,036 images and 5,320 labeled instances. Both collections encompass a wide range of complex mining landscapes characterized by diverse shapes, contextual settings, and color attributes, including difficult examples that exhibit strong spatial and spectral resemblance to adjacent terrain. These resources are designed with attention to land use strategy and ecological governance, aiming to advance the systematic implementation of automated surveillance and management for open-pit mining areas.