Use this URL to cite or link to this record in EThOS:
Title: Attitude determination by exploiting geometric distortions in stereo Earth images
Author: Mumtaz, Rafia
ISNI:       0000 0004 2684 587X
Awarding Body: University of Surrey
Current Institution: University of Surrey
Date of Award: 2009
Availability of Full Text:
Access from EThOS:
Access from Institution:
The growing interest in the development of small satellites and the demand for high resolution imaging has made the pointing and drift rate requirements of a satellite more stringent. To achieve high pointing accuracy star sensors can be used but their cost are too large for small satellites. The need for keeping the overall cost of the spacecraft down and still achieve adequate pointing accuracies has provoked the development of relatively inexpensive and high performance attitude systems that can provide competitive pointing accuracies during imaging operations. In order to realize such an attitude system, this research has exploited the offsets present between pairs of ground images for determining the orientation of the spacecraft. The approach is based on a pair of canted Earth pointing pushbroom sensor. The configuration of the sensors induces temporal and perspective distortions in the images. These geometric distortions give rise to interimage offsets. A mathematical model has been developed to establish a relationship between the attitude and the image offsets. The mathematical modelling of the attitude system is based on the camera system of DMC satellites. However, this model is also applicable to other satellites with similar sensor configuration. To evaluate the performance of the attitude model with images, simulated satellite images have been created with known ground truth. These simulated images have been generated by an extension of the attitude model. The offsets between the images have been found in Fourier domain using Singular Value Decomposition (SVD) factorization. To determine the attitude from the image shifts, the attitude model has been inverted. In order to quantify the model with images, attitude has been estimated from a number of synthetic images generated with an angular separation, α, of 0.045° between the sensors. It has been found that with an α of 0.045° between the sensors, this algorithm gives an accuracy of ±0.16°(1σ). However, increasing the angular separation to 0.5°, the accuracy of 0.034°(1σ) could be attained. The attitude model has also been tested with different experimental images of DMC satellites and the results are compared with the onboard ADCS data. The onboard vibrations are also determined for UK DMC, ALSat-1 and Nigeria Sat. The experimental results have shown the potential of the image based attitude model to complement the need for low cost, low mass, and highly accurate attitude system for small satellites.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available