Desert Convoy Crossing Route Video Dataset

In vast desert areas, road traffic and environmental monitoring have always been significant challenges. Traditional manual monitoring methods are inefficient and dangerous, and the lack of precisely annotated data for automated systems poses a dilemma for path planning and environment recognition by small unmanned convoys. This dataset aims to provide high-quality video data for autonomous driving and environmental monitoring systems to better tackle the complex and variable desert environment. Data is collected by HD cameras installed at the front of convoy vehicles, selecting covered path videos under different weather and lighting conditions. Multi-round annotation and expert review are conducted for data quality control, operated by a professional team with backgrounds in traffic and environmental science. Data is stored in MP4 format to ensure high compatibility and easy accessibility. Preprocessing includes video clipping, background noise cleaning, and scene change detection technical steps.