Farmland Type Remote Sensing Classification Dataset

The current agricultural industry faces issues with crop monitoring and management efficiency, especially in large-scale farmland monitoring and data analysis. Traditional methods are often time-consuming and costly. Although existing remote sensing technology has been applied to agriculture, the lack of accuracy and reliability in data classification results in suboptimal decision support systems. This dataset aims to solve the technical challenge of crop type recognition by providing high-quality farmland remote sensing images, meeting the practical needs of agricultural production and environmental monitoring. The data is captured by drones in different seasons and locations, ensuring coverage of diverse crop types and growth conditions. To ensure data quality, we adopted a multi-round annotation and expert review approach to guarantee consistency and accuracy of annotations. The data is stored in JPEG format, organized by time and location for convenient subsequent data analysis and model training.