Accurate extraction of the current characteristics of check dam is very important for the evaluation of check dam, and the comprehensive characteristic information of check dam can provide an important basis for the reinforcement of diseased and dangerous dam. However, the current check dam feature information extraction still lacks scientific and efficient technology. To address this problem, the proposed method was based on the unmanned aerial vehicle (UAV) tilt photography technology to extract the characteristics of the current status of check dam operation. UAV aerial images were used as the base data to generat a 3D real scene model and a high-precision DEM for the extraction and measurement of important current information such as the number, location, remaining storage capacity, and effective dam height of the check dam. and then analyze the distribution, composition structure, degree of siltation, dam-controlled watershed area and other characteristics of siltation dams in a small watershed. This technology was applied to the small watershed of Shejiagou in Zizhou County, Yulin City, Shaanxi Province, to extract detailed information on the current status characteristics of one large-scale dam and three mediumsized dams. After comparing with the measured data in the field, the relative error of the check dam body information was less than 3.8%. The identified large dams had insufficient remaining capacity and were in a largely silt-filled state, and the crest has obvious subsidence, which has a significant risk of dam failure under extreme precipitation conditions. The proportion of the dam land cultivated area to the small watershed area was only 2.17%, there was still land uncultivated as farmland. The research reaults can provide a new method and idea for dynamic monitoring and scientific management of check dams in small watersheds.