As a critical soil and water conservation project in the Loess Plateau region, check dams play a central role in controlling soil erosion and ensuring regional food security. However, their intelligent management has long been constrained by technical bottlenecks such as inefficient data acquisition, insufficient model generalization capabilities, and lack of standardized datasets. This study utilized 0.75-meter high-resolution Jilin-1 remote sensing imagery and selected the Jiuyuangou watershed, a typical watershed in the Loess Plateau. Through a systematic sample preparation workflow, including grid division, sample screening, pixel-level semantic annotation, and image enhancement, the dataset comprises 2,920 high-precision samples that comprehensively ensure the completeness of spatial representation and satisfy the requirements for algorithm generalization. Data quality evaluation experiments demonstrated that models trained on this dataset (e.g., DenseUnet) achieved excellent performance on the validation set, with a mean Intersection over Union (mIoU) exceeding 80% and Overall Accuracy (OA) reaching over 89%. Compared with public datasets, the spatial consistency and reliability of silted land formed by check dam extraction results were significantly improved. This dataset enables fine-grained semantic differentiation between check dam sedimentation land structures and surrounding backgrounds under complex geomorphological conditions, addressing the lack of standardized datasets for intelligent identification of silted land formed by check dams. It not only provides crucial data support for high-precision spatiotemporal mapping of silted land formed by check dams, dynamic assessment of dam-break risks, and quantitative analysis of soil conservation effects, but also opens new technical pathways for AI-driven optimization of soil and water conservation engineering. These advancements hold substantial practical value for promoting ecological protection and high-quality development in the Yellow River Basin.
LIU et al. (2026) studied this question.