Image based integrated satellite attitude determination

In this thesis, an image-based attitude determination system, which determines the attitude of a satellite on three axes with images of the Earth and the Sun by using one or two cameras in the satellite body and a simulation environment based on computer vision were presented. In addition, a gyroscope integrated satellite attitude determination system, which uses the image based attitude determination system outputs as a reference solution has been proposed. TRIAD method was used for image-based attitude determination. The results of the simulation showed similar results in determining image-based attitude for single-camera and two-camera architectures; however, the solution could not be produced in a single camera system when the Earth and the Sun were not on the image plane at the same time. The average angle error on each axis of the image-based system in the medium Earth orbit and geosynchronous orbit was detected below 1 degree, however the system alone could not achieve sufficient results for low Earth orbit. It has been observed from the simulation results of the integrated system that the attitude information can be derived with the gyroscope rate data for short periods of time in the case where the image-based attitude solution can not be produced. By achieving image-based attitude solution, gyroscope rate errors can be corrected and the growth of the integrated attitude determination system error can be prevented.