SBIR Phase I: Autonomous 3D Scanner for Building Interiors and Exteriors

Period of Performance: 01/01/2013 - 12/31/2013


Phase 1 SBIR

Recipient Firm

Scientific Imaging and Visualization, LLC
1996 Wellesley Trace
Atlanta, GA 30338
Principal Investigator, Firm POC


The innovation of this project is the development of a system that can autonomously scan building interiors and exteriors and construct corresponding detailed, photo-realistic 3D models with minimal interaction of an operator that does not need to have specialized skills. The idea is to use a small unmanned aerial vehicle (UAV) equipped with a wireless camera and a laser line that can almost fully autonomously fly inside and around the building, video record all the interior and exterior structures, and send the video and other recorded information to the base station where the 3D model is constructed using computer vision techniques. The proposed system requires the development of two innovative technologies: (1) automated construction of texture-mapped 3D models of structures video recorded from a flying vehicle, and (2) autonomous flying of an UAV inside and around a building based on the recorded video and range information. While 3D scanning is an existing technology and small UAVs have already been developed, this proposal will combine these technologies and enhance their capabilities to deliver new value. In addition, computer, smartphone and tablet software for interactive visualization and exploration of the constructed 3D models will be developed. The broader/commercial impact of this project is in the area of marketing of businesses and institutions, virtual reality applications as well as in rescue and law-enforcement applications. Having detailed, photo-realistic 3D models of building interiors and exteriors is useful in a number of applications. However, the existing approaches for the construction of such models are labor intensive and require highly trained people and expensive equipment, which make the entire process prohibitively expensive for most applications. The proposed system would reduce the cost of construction of such models by at least an order of magnitude compared to the existing solutions and would automate most of the process. The proposed system could be used to relatively inexpensively construct detailed, photo-realistic 3D models of interiors and exteriors of: real estate properties, restaurants, sport clubs, museums, historic buildings, architectural sites, touristic sites, university campuses, amusement parks, water parks, zoos, apartment complexes, resorts, hotels, hospitals, shopping malls, etc. The constructed 3D models can then be explored online using computers, smartphones and tablets. Furthermore, the system can be used for mapping or exploring buildings in situations where it is dangerous or impractical for people to enter, which is often the case in rescue and law-enforcement applications.