Date of Award

Spring 2026

Access Type

Thesis - Open Access

Degree Name

Master of Science in Aerospace Engineering

Department

Aerospace Engineering

Committee Chair

Thomas Alan Lovell

Committee Chair Email

lovel7e5@erau.edu

First Committee Member

Thomas Alan Lovell

First Committee Member Email

lovel7e5@erau.edu

Second Committee Member

Morad Nazari

Second Committee Member Email

nazarim@erau.edu

Third Committee Member

Troy Henderson

Third Committee Member Email

hendert5@erau.edu

College Dean

James W. Gregory

Abstract

One of the fundamental tenets of Space Situational Awareness (SSA) is the detection and sub sequent identification of Resident Space Objects (RSOs) within unresolved optical space imagery. This function is vital to the documentation and tracking of RSOs in their respective operational orbits, knowledge that is necessary for collision avoidance efforts and Space Domain Awareness (SDA) applications. In previous work, development was begun on a MATLAB program called RSOID to fulfill this purpose by accepting a collection (or ’collect’) of unresolved imagery and outputting a sequence of RSO locations (called a ’tracklet’) that can be used to determine the RSO’s orbit or correlate it with an existing RSO catalog. With the advent of the SSA field’s rapidly increasing popularity, extensive research is being conducted on various ways to streamline and optimize SSA missions. One of these objectives is autonomous SSA activity, which requires automated RSO identification capability. The RSOID program relies on multiple parameters that heretofore were arbitrarily defined by the user. Therefore, in the pursuit of autonomous SSA activity, this research focuses on the development of adaptive methods for some of those parameters, particularly those of the Star elimination step of the RSOID algorithm. Star elimination is a process by which the detected image objects belonging to stars are identified and removed such that the RSO tracklet becomes more isolated, making the RSO location data easier to determine and interpret. Some of the outcomes of this research are methodologies of autonomously selecting the parameters that Star elimination relies on. Furthermore, the Star elimination process as a whole was restructured to accommodate the adaptive methods and implement perceived improvements during the course of the research. The research also lead to the development of a type of Star elimination that was new to the RSOID project called star catalog comparison, which makes use of a reference star catalog to identify and eliminate stars; whereas previously, Star elimination relied solely on the data gleaned from the detected objects within the imagery alone to identify stars. This new type of Star elimination introduces a new set of parameters to define, prompting the development of schemes to adaptively select those as well. Both the reworked Star elimination method and the new catalog comparison method prove to be more robust and effective at eliminating detected stars compared to the preexisting version; however, this comes at the cost of increased runtime. Recommendations for further research are the optimization of these new methods of Star elimination to reduce computational cost, as well as the development of adaptive schemes for other user-defined parameters of the RSOID program, as the progression toward autonomous SSA capability continues.

Download

Included in

Astrodynamics Commons

Share

COinS