3D Model Reconstruction from 3D Point Clouds Generated from Digital Images

BIM applications, traffic control, surface management with visualization, and the restoration of historic structures all need the creation of 3D models. However, laser scanning equipment with a lidar sensor is expensive, and lower-accuracy devices are cheaper than expensive ones. This research paper aims to reconstruct a 3D information model from photographic images to save money and be more flexible than terrestrial laser scanning, ensuring coverage and satisfying the required error condition. Furthermore, it can build point clouds containing red, green, and blue (RGB) values, which is impossible with low-cost lasers. The 3D model reconstruction approach was used in this study to work with digital photos. Creating the point cloud model from the acquired images is possible using Structure from Motion. The Mesh model is then created using the Revit Architecture software based on the generated point cloud. The final step is to analyze the divergence, make remarks, and applicability of the approach by comparing and contrasting the direct measurement method of the item with the point cloud model and 3D information model.