Grant: Collaborative Research and Development Grants

Personal identification no. (PIN) Family name of applicant Jenkin SUMMARY OF PROPOSAL FOR PUBLIC RELEASE (Use plain language.) This plain language summary will be available to the public If your proposal is funded. Although tt is not mandatory, you may choose to The ability to construct high fidelity complex surface models of structures using 3d sensors is an enabling technology. When coupled with appropriate algorithms to merge together models obtained over larger geographical extents these sensors can be used to build complex models of structures from archaeological treasures to crime scenes and from industrial parts to sculpture. Although these sensors and their corresponding algorithms can produce extremely large complex models the models that they produce are not always directly usable by other agents. They contain very little semantic information about the structures scanned beyond physical surface properties. Yet it is the semantic properties of these scans the foreground object, specific components of the object ~ that are often the reason for the object being scanned in the first place. How can these scans be manipulated so as to segment the salient components of the scans? How can these components be represented so that queries of the salient components of these scans can be handled efficiently and effectively? How can this information be used in a timely fashion in order to guide the process of scan acquisition? Building upon an existing research collaboration between MDA and York University, this project investigates issues related to the segmentation, representation, retrieval and manipulation of semantic objects from 3D sensor scans of extended physical environments. Second Language Version of Summary (optional).