Font Size: 

Underground space is an important natural resource for urban development. The assessment of the exploitable capacity of urban underground space is conducive to understanding the total quantity and potential of urban underground resources that can be developed and utilized. However, there is currently a lack of automatic and systematic evaluation methods of urban underground space resources capacity. In this paper, an urban underground space resource capacity evaluation method based on borehole data, geological profiles, resource development survey results and satellite images are proposed. As the basis of the evaluation frame work, a 3D geological model is established using the borehole data and geological profiles, and the ordinary Kriging interpolation method is applied to attribute interpolation for the establishment of the attribute model. After the property model was established, we also integrated the resource development survey results with the model. In addition, the influence of existing buildings on underground space development is estimated automatically from satellite images through image segmentation algorithm based on deep learning methods. The proposed method is used to evaluate the underground resource capacity in the Sanlong Bay area, Foshan City, Guangdong, China. The achieved result provides an effective and automatic way of estimating the underground capacity and is conducive to urban planning and underground constructions.