3D Modeling for Indoor Structure Using Omniverse Create

Following the increasing demand for 3D reconstruction of indoor structures, this study focuses on the current state of 3D mapping and modeling software available. However, reconstructing 3D models of indoor environments is more challenging due to spatial constraints and the complexity of objects within them. By using an application known as Polycam, a 3D map of an indoor building is generated. Polycam has a feature called the Photo Mode that enables the user to capture multiple sets of images of an object or surrounding to create a 3D model from the data acquired. It will automatically stitch the image together to create a single 3D model of the object. By utilizing this feature, several images of a room are captured for data acquisition. Additionally, NVIDIA Instant-NGP is also utilized for data acquisition which uses the NeRF technology and COLMAP to extract a video input into several frames or images which can then be trained to obtain the 3D data as a whole. The data acquired are then exported to a format that is supported by Omniverse Create to complete the 3D reconstruction. Omniverse Create is 3D modeling software that can be said which is relatively new to the 3D modeling scene and it is able to simulate real-time lighting situation. Nevertheless, the outcomes of the 3D mapping process failed to align with the initial anticipations as a result of various constraints and limitations.