Abstract

Activity Based Intelligence (ABI) is the derivation of information from a series of in- dividual actions, interactions, and transactions being recorded over a period of time. This usually occurs in Motion imagery and/or Full Motion Video. Due to the growth of unmanned aerial systems technology and the preponderance of mobile video devices, more interest has developed in analyzing people's actions and interactions in these video streams. Currently only visually subjective quality metrics exist for determining the utility of these data in detecting specific activities. One common misconception is that ABI boils down to a simple resolution problem; more pixels and higher frame rates are better. Increasing resolution simply provides more data, not necessary more informa- tion. As part of this research, an experiment was designed and performed to address this assumption. Nine sensors consisting of four modalities were place on top of the Chester F. Carlson Center for Imaging Science in order to record a group of participants executing a scripted set of activities. The multimodal characteristics include data from the visible, long-wave infrared, multispectral, and polarimetric regimes. The activities the participants were scripted to cover a wide range of spatial and temporal interactions (i.e. walking, jogging, and a group sporting event). As with any large data acquisition, only a subset of this data was analyzed for this research. Specifically, a walking object exchange scenario and simulated RPG. In order to analyze this data, several steps of preparation occurred. The data were spatially and temporally registered; the individual modalities were fused; a tracking algorithm was implemented, and an activity detection algorithm was applied. To develop a performance assessment for these activities a series of spatial and temporal degradations were performed. Upon completion of this work, the ground truth ABI dataset will be released to the community for further analysis.

Library of Congress Subject Headings

Video surveillance--Data processing; Automatic tracking--Data processing; Computer vision

Publication Date

8-15-2014

Document Type

Thesis

Student Type

Graduate

Degree Name

Imaging Science (MS)

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

David Messinger

Advisor/Committee Member

Carl Salvaggio

Advisor/Committee Member

Derek Walvoord

Comments

Physical copy available from RIT's Wallace Library at TA1637 .L494 2014

Campus

RIT – Main Campus

Plan Codes

IMGS-MS

Share

COinS