 |
General: 663 |
P5: Generating Compressed Point Cloud Image Representations
Author: Christopher Miller, Advisor: John Benedetto (NWC, Mathematics)
Problem Statement Presentation
Project Proposal
Abstract
Urban terrain data can be conceptualized as a piecewise two-dimensional manifold embedded in a compact subset of R3. A problem arising frequently in image analysis is, given a discrete set of points sampled from a manifold called a point cloud, how can one form an approximation of the manifold. We propose the development of software suited to this task when the point cloud is formed by an aircraft using a Light Detection an Ranging (LIDAR) system to create a point cloud from a section of urban terrain. the key advantage of this algorithm over others is that the image representation will be highly compressed. This will allow data acquisition, transmission, and use to occur in real time.
MidYear Progress Report and Presentation
Final Presentation , Final Report
