To tackle the difficulty of topographic detection inside caisson during construction monitoring, we put forward a detection method for deep-water caisson in closed environment based on analyzing the operation environment. The interior topographic data of caisson was successfully captured by 3D multi-beam real time sonar system in the Oujiang Beikou Bridge. The 3D scene data of caisson’s interior topography were constructed by GIS technology. Moreover, according to rope survey data and soil excavation scheme, the accuracy of the detected topographic elevation is verified to meet the construction requirements, and the expression of surface morphology is proved credible, which indicate that the proposed method is feasible. By analyzing the detection data we found that different soil excavation zones can be identified by the brightness and shade of the detection data image with the spatial geometric morphological characteristics. Visual analysis of soil excavation zones and topographic elevation in the assistance of GIS technology will provide auxiliary data for determining the approach, position, and range of soil excavation for caisson construction.