Masters Theses
Keywords and Phrases
active space sustainability; astrodynamics; pose estimation; space domain awareness; space sustainability
Abstract
"Spacecraft rendezvous is an integral part of many orbital missions, including active space debris removal (ADR) and in-orbit servicing, assembly, and manufacturing (ISAM). Traditionally, rendezvous missions rely mainly on manual ground-based control to perform rendezvous and docking. While this can be effective on an individual basis, it becomes less practical at scale. Autonomous vision-based methods can reduce reliance on human control, and may improve scalability, response time, and reduce the possibility of human error for both ADR and ISAM. This work lays the necessary groundwork for developing a machine learning (ML) model for autonomous rendezvous by creating a labeled image dataset that could be used for training and evaluation. Due to the impracticality of collecting labeled images in orbit, alternative methods are used to simulate the space imaging environment, both through computer-generated synthetic images, and laboratory-based simulated images. A 3D environment is built in Blender and used to generate 62,482 synthetic images for training, validation, and testing. For simulated images, the work covers the development of the laboratory setup, calibration process, and initial dataset generation framework. Together, these contributions provide a foundation for future vision-based pose estimation research by the DCV-Space group"-- Abstract, p. iii
Advisor(s)
Nandan, Smriti (Paul)
Committee Member(s)
Han, Daoru Frank
Pernicka, Henry J.
Department(s)
Mechanical and Aerospace Engineering
Degree Name
M.S. in Aerospace Engineering
Publisher
Missouri University of Science and Technology
Publication Date
2026
Pagination
vii, 43 pages
Note about bibliography
Includes_bibliographical_references_(pages 41-42)
Rights
© 2026 Mohanad Alameer , All Rights Reserved
Document Type
Thesis - Open Access
File Type
text
Language
English
Thesis Number
T 12612
Recommended Citation
Alameer, Mohanad, "Aura-Visionpose: Creating Datasets for Vision-Based Pose Estimation" (2026). Masters Theses. 8290.
https://scholarsmine.mst.edu/masters_theses/8290
