SBIR-STTR Award

Freestyle Technology, Inc. proses to design an build a six-camera video imaging device for use in navigation research. The individual cameras will be attached to two-meter tethers to allow for flexible placement on a variety of platforms. All cameras will be genlocked to a common signal to assure that corresponding pixels in each camera are sampled at the same time. A user interface will be provided to allow the user to set individual camera parameters and to sample individual frames or sequences of frames from the camera system. Optical flow algorithms will be developed and embdedded in the user interface to allow for the estimate of motion parameters of interest for navigation purposes. A multiple-view genlock camera (MCGV) system will be of value for all types of robotic air and land vehicles. It is anticipated that a navigator based on the MVGC will be of lower cost than a MEMS-based IMU and have a lower drift rate.

Keywords:
Multiple-View Camera, Optical Flow , Navigation

The multiple-view genlocked camera (MVGC) developed and demonstrated duringthe first phase of this program will be extened to the long infrared domain. In addition, real-time structure from motion (SFM) and video-based anvigation algorithms will be coded on the Phase I and Phase II ontrol computers. Captive Flight Tests in winter and summer conditions will be executed to gather data and to evaluate perfomance of the cameras and algorithms.

Keywords:
Multiple-View Cameras, Multiple-View Geometry, Structure From Motion, Optical Odometry