Open Competition 3 - Information Technology
Integration of 3-D Photography for Photorealistic Modeling in Construction and Disaster Recovery
Develop software that creates accurate, realistic, three-dimensional models of buildings and other large-scale environments for applications in construction, disaster recovery, architecture, urban planning, preservation, and entertainment.
Sponsor: BRAINSTORM Technology LLC514 West 24th St.
New York, NY 10011
Brainstorm will develop affordable computer software that enables accurate, highly realistic three-dimensional (3-D) photographic modeling on a large scale so that rescue and emergency workers can thoroughly acquaint themselves with specific buildings and other major structures and plan responses in the event of a fire or other disaster. High-quality 3-D models of real-world objects as large as New York City's Grand Central Station would be created by fusing images from digital cameras with data from advanced range scanners. Current 3-D photographic computer modeling is technically inadequate and very expensive. Unlike two-dimensional images acquired through digital photography, 3-D models require painstaking specification that is labor intensive and very costly. Yet, the resulting models lack the precision necessary for many applications. With the technology it is pursuing, Brainstorm believes the complex task of constructing 3-D site models should be reduced from months and hundreds of man-hours to just hours of computation with minimal human interaction. The recent emergence of high-resolution range scanners that can capture details as small as 6 millimeters from a distance of 50 meters may open the way to a solution to this problem. However, the voluminous data produced by these scanners and digital cameras would be overwhelming without an integrated software system of the type Brainstorm aims to develop. The three-year ATP project will require considerable innovation to overcome technical obstacles in 3-D segmentation, registration, texture mapping, 3-D modeling, and visualization. Major technical risks stem mostly from problems of scale--mathematical computations to derive and visualize accurate three-dimensional large-scale models are dauntingly complex, while data processing and storage requirements are enormous. The initial goal is to create a manageable interactive system that saves time and money while developing models that demonstrably improve safety. ATP funding is needed to enable Brainstorm to make the technological leap needed to develop 3-D photography software that forms the basis for photorealistic modeling. Brainstorm is a small company that does not have the ability to fund this three-year project, and several attempts to obtain venture capital funding were unsuccessful. Other potential application areas for this technology include: entertainment, engineering, surveying, military and urban planning, archaeology, and historic preservation. A successful software solution for large-scale 3-D modeling could lead to many new markets within industries that could not previously afford 3-D modeling. Eventually, as range scanners become more widespread, Brainstorm plans to provide the service on-line as an application service provider, further enabling broad diffusion of the benefits.