Aligning Astronomical Telescopes via Identification of Stars
- Friday, 01 October 2010
The alignment process would be completely automatic.
A proposed method of automated, precise alignment of a ground-based astronomical telescope would eliminate the need for initial manual alignment. The method, based on automated identification of known stars and other celestial objects in the telescope field of view, would also eliminate the need for an initial estimate of the aiming direction. The method does not require any equipment other than a digital imaging device such as a charge-coupled-device digital imaging camera and control computers of the telescope and camera, all of which are standard components in professional astronomical telescope systems and in high-end amateur astronomical telescope systems. The method could be implemented in software running in the telescope or camera control computer or in an external computer communicating with the telescope pointing mount and camera control computers.
The image in the telescope field of view would be captured by the digital imaging device and digitized and then, according to the method, would be processed by a variant of any of several previously published star-identification algorithms. In simplified terms, such an algorithm determines criteria such as brightnesses and relative angles or distances between stars in the digital image and matches those criteria with stars in a database. Once such a match was found, the celestial coordinates of the identified objects in the image and the pixel coordinates of the object would be used to precisely determine the line of sight of the telescope in celestial coordinates.
Although the method does not require an initial estimate of the aiming direction, such an estimate (or ancillary information from which such an estimate can be calculated) could be used to accelerate the automated precise alignment process by limiting the search space to a small portion of the celestial-object database. Even if all that is known are the geographic coordinates of the telescope and the time, portions of the sky known not to be visible from that location at that time could be excluded from the search.
Once the celestial coordinates of two different lines of sight have been deter- mined precisely as outlined above, the telescope would be automatically initial- ized and aligned for subsequent auto- mated pointing and tracking. Thereafter, during tracking, the align- ment process as described thus far could be repeated as often as desired to update the alignment: At each update, the celestial coordinates of the current line of sight would be communicated to the telescope control computer to main- tain or restore the precise alignment of the telescope drive axes. Because the line-of-sight directions determined by this method would be based on direct observation of celestial objects having known coordinates, they would be more accurate than are the line-of-sight direc- tions determined by prior methods that involve intermediate measurements (e.g., drive-shaft-angle measurements), which introduce drive-train and axis-mis- alignment errors.
This work was done by Mark Whorton of Marshall Space Flight Center.