Reconstructing viewing parameters from images as a tool in airplane assembly

During assembly of an aircraft, technicians working in the interior need to be able to quickly identify parts and access part numbers. Part numbers can be retrieved through 3D geometric models of the aircraft. Navigating large 3D models is challenging and time consuming, and it is nearly impossible on small mobile devices with limited processing power and screen sizes. We will explore the use of techniques from computer vision and heuristic search with the goal of eliminating the need to manually traverse large 3D geometric models or search long parts catalogues. Technicians use digital cameras to take a picture of the area in question. That picture is matched with the 3D model of the aircraft and the viewing parameters (camera location and orientation) are determined. From those, a view of the 3D model that correctly reflects the technician's location and view is generated. Part numbers and associated information can be retrieved easily.

Faculty Supervisor:

Dr. Dirk Arnold

Student:

Homa Foroughi

Partner:

AeroInfo Systems - A Boeing Company

Discipline:

Computer science

Sector:

Aerospace and defense

University:

Dalhousie University