Model-based recognition and range imaging for a guided vehicle

摘要

Progress in providing the Oxford Autonomous Guided Vehicle (AGV) with range sensing and model-based recognition of classes of industrial artifacts is described. The GEC ‘Turtle’ vehicle is the basis for a sensor integration project in which vision sensors play the strategic role of locating pallets in an unstructured environment; we report on two key areas involved with achieving this goal. The design and geometry of a novel range sensor developed jointly with the National Engineering Laboratories is described. The problems of dealing with noisy range images are discussed, and a number of solutions proposed and tested. A system for pose recognition of generic models from range-imagery is described.