Student: Nykia L. Jackson
Senior Researcher: Dr. Craig Scott
Title: Database Integrity Monitoring of Synthetic Vision Systems using Video Mosaics
In synthetic vision systems, there exists a need to develop a methodology to ensure process accuracy and database integrity during the lifecycle of the actual database. Preliminary research in the on-going development of Database Integrity Monitoring Equipment (DIME) is also focused on real-time accuracy and integrity work. To achieve primary flight display certification, the DIME must provide adequate integrity, reliability and accuracy. This is a tedious and time-consuming technique, requiring computation of several mathematical equations and algorithms. This research aims to improve the DIME by using stereo 3D mosaics of video sequences captured by the aircraft’s onboard sensors to compute real-time comparisons between the stored database and actual sensor terrain measurements.
Video mosaicing in conjunction with reconstruction algorithms will be used to acquire the image and reconstruct its spatial data. Several algorithms will be tested to compute the critical parameters including corner detection, correlation matching, and the fundamental matrix. The depth map will be generated using the most efficient algorithm with the best resolution and compared to a digital elevation model (DEM) file acquired from the USGS database.
In real-time, implementation of this stereo reconstruction algorithm into the DIME may enable pilots to perform database integrity verification, while increasing situational awareness and improving the aircraft’s overall safety. Future work will include performing error analysis, developing a fully functional prototype where speed is negligible and finally integrating the algorithms into hardware to increase the database verification speed to near real-time.
|
