Urban Low-Altitude Drone Vehicle Image Dataset

As urban traffic becomes increasingly congested and the number of vehicles dramatically increases, traffic management faces enormous challenges. Traditional ground monitoring systems are limited in terms of perspective and range, making it difficult to provide comprehensive and effective data support, while the rise of drone technology offers new solutions for traffic monitoring. However, existing datasets mostly rely on fixed camera equipment and lack vehicle image data support from dynamic perspectives, making them unable to meet current technological needs. The construction of this dataset aims to provide high-quality vehicle images captured from urban low-altitude drone aerial photography, addressing the challenges of vehicle detection in dynamic environments. The data is collected through high-resolution drone equipment over multiple periods and different weather conditions in urban areas, ensuring diversity and representativeness of the data. Strict quality control employs a strategy of multiple rounds of annotation combined with expert review to ensure consistency and accuracy of annotations. The annotation team consists of over 30 traffic field experts and image processing professionals. Data preprocessing includes carefully reviewed image enhancement and format conversion, ultimately stored in JPG format with unified organization. Vehicle information is clearly categorized by different classes, making it easy to retrieve and use.