Automated Bottom-Up, Top-Down Method for Creating Digital Staircase Models from Dense Point Clouds Using Parametric Prototype Models and Model Fitting

Digital building models have become a transformative tool in the management and monitoring of assets in the built world. Laser scanners are valuable tools for creating digital building models through Scan-to-BIM algorithms, capturing the as-built status of the environment in the form of point clouds. However, a main challenge persists in the automated creation of digital models from these point clouds, providing both coherent geometry, and semantics. Staircase elements are vital in multilevel buildings, facilitating essential vertical movement and serving as crucial emergency evacuation routes. This paper proposes a hybrid bottom-up, top-down approach for the automatic creation of digital staircase models using laser scanner point clouds. The workflow involves separating staircase points, designing parametric models, and model fitting through optimization. The proposed method is validated using eight real laser scanning point clouds containing highly diverse stair configurations. The results demonstrate the effectiveness of the proposed method in automatically creating high-quality digital staircase models with coherent geometry.