SBIR-STTR Award

Currently there are no automated and efficient portable 3D data acquisition methods to obtain accurate and detailed geometry and texture data for use in computer-based simulators. The goal of the proposed project is to develop innovative methods of data acquisition and processing to produce a quick, accurate, and portable 3D data acquisition system to capture geometry and texture data. The system will use automated video processing, accompanies by new uses of positioning devices, and photogrammetric triangulation. In addition to unique applications of state-of-the-art hardware, such as the use of motion video, new software will be developed to automate the measurement process and obtain the detail and realism required for simulation. Recorded data will be usable by existing geometry databases, such as NPSNET, to allow better simulation on existing and new military systems. Specific areas of development will be photographic methods, positioning devices, data processing methods, geometry calculation procedure, and interfacing with databases. Emphasis will be placed on development of a system which will require the least compromise on speed and realism in acquiring, processing, and displaying geometries and textures. CamSys, Inc. has extensive experience in producing and marketing automated measurement equipment for acquiring geometrical data, and has distinct capabilities needed to develop a system for the Naval Training Systems Center. The proposed work examines the feasibility of such a system for Phase II development of a prototype.

Accurate and cost effective visual databases, containing geometry and texture data, are needed in themilitary and civilian secotrs to produce a realistic description of specific surroundings for training simulators. Currently, geometry data for visual databases are commonly generated by creating three-dimensional models from photographs and terrain elevation maps using modeling software. Present methods to obtain texture data for visual databases rely heavily on land use maps and satellite data. However, the visual data produced by these labor intensive methods lack realism, resulting in noticeable differenced between the computer generated display and the real images. The goal of this proposed Phase II project is to produce the prototype of a portable, easy-to-use, and automated system to extract geometry and texture data from video images of real surroundings using the new analytical methods developed during the Phase I project. To acquire the necessary data, the system will use a camcorder, Global Positioning System (GPS), electronic compass, and laptop computer. Video images will be processed for geometries and textures using methods to identify two-dimensional points of user specified geometries, calculate the three-dimensional points, and extract textures. Visual databases containing geometry and texture data will be further developed for military and commercial applications.