Spatial and visual data fusion for capturing, retrieval, and modeling of as-built building geometry and features

Abstract

Abstract Background As-built building information, including building geometry and features, is useful in multiple building assessment and management tasks. However, the current process for capturing, retrieving, and modeling such information is labor-intensive and time-consuming. Methods In order to address these issues, this paper investigates the potentials of fusing visual and spatial data for automatically capturing, retrieving, and modeling as-built building geometry and features. An overall fusion-based framework has been proposed. Under the framework, pairs of 3D point clouds are progressively registered through the RGB-D (Red, Green, Blue plus Depth) mapping. Meanwhile, building elements are recognized based on their visual patterns. The recognition results can be used to label the 3D points, which could facilitate the modeling of building elements. Results So far, two pilot studies have been performed. The results show that a high degree of automation could be achieved for the registration of building scenes captured from different scans and the recognition of building elements with the proposed framework. Conclusions The fusion of spatial and visual data could significantly facilitate the current process of retrieving, modeling, and visualizing as-built information.