Farmland and Woodland Boundary Detection Dataset

The current agricultural sector faces the dual challenge of increasingly scarce land resources and environmental protection, making the detection of farmland and woodland boundaries a critical issue that needs urgent resolution. Existing monitoring methods often rely on manual surveys, which are inefficient and prone to errors, failing to meet the needs for speed and accuracy. Hence, we have developed this dataset to provide high-quality training data for automated boundary detection, meeting the technological needs of precision agriculture and sustainable development. The dataset is obtained through drone aerial photography, capturing diverse farmland and woodland environments, covering different seasons and climatic conditions. To ensure data quality, we employ multi-round annotation and consistency checking mechanisms, ensuring each image is reviewed by experts with accurate annotations. Data is stored in JPG format, organized as needed, facilitating subsequent deep learning model training and validation.