Abstract

The Digital Imaging and Remote Sensing Image Generation (DIRSIG) model and other image simulators provide the ability to utilize detailed, artificial scenes to generate spectrally and spatially realistic simulated imagery. Simulated imagery is useful in a myriad of ways, such as sensor modeling, algorithm performance assessment, and others. Actually making synthetic scenes, however, is often a time consuming process, requiring the manual placement of the many objects required to define the scene. This is particularly true for scenes of large spatial extent. Proposed is a technique to generate large-area night scenes for DIRSIG. This is accomplished by using freely available Geographic Information System (GIS) data to inform the placement of street light sources. Results to this point have demonstrated that this technique is a feasible way to model the radiance for large urban areas. This determination was made through comparison to real night time data collected by the Visible Infrared Imaging Radiometer Suite (VIIRS). The methodology is presented as a modular framework, so that future researchers can recreate the work done do this point, with the ability to easily substitute components of the workflow, such as using an alternate source of GIS data or a different simulation environment.

Library of Congress Subject Headings

Geographic information systems--Mathematical models; Geographic information systems--Data processing; Optical images--Computer simulation

Publication Date

8-14-2012

Document Type

Thesis

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

Messinger, David

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: G70.23 .Z65 2012

Campus

RIT – Main Campus

Download

Share

COinS