DSpace at Bangkok University >
Graduate School >
Master Degree >
Theses >

Please use this identifier to cite or link to this item: http://dspace.bu.ac.th/jspui/handle/123456789/3912

Authors: Gaël Robin
Keywords: Digital refractor telescope
Raspberry pi application
image processing
Raspberry pi application
cameras calibration
Issue Date: 26-Jun-2019
Abstract: The on-going development of image acquisition tools makes possible the night sky studies by amateurs, students and professional. Digital telescopes are available but it comes to a high-cost. The project presented aimed at providing education institutions in developing countries and amateurs a low- cost digital optical telescope. This telescope is 3D printed based structure and it is controlled remotely through the Internet with an in-house developed software with the use of Java. The motorization of the telescope is ensured by two steppers motors coupled with pulley belts and gears in order to decrease the rotation angle to improve the motion accuracy. The control is performed using a Raspberry Pi 3 to ensure the connection between the client and the telescope. A MQTT server, a lightweight communication protocol known in the Internet of Things field, is used to provide a full communication between the telescope and the in-house control-software. The telescope itself is composed of two web-cameras: a large field of view for global localization, and another one coupled with an achromatic lens proving a high magnified image for detailed observations of celestial objects. The telescope structure is made of 3D printed PLA parts, which keeps it low-cost and robust. The video feeds from the two web-cameras are displayed on the control software that provides the possibility to save snapshots. Also, in order to compensate the shift between the finder-scope center point and the telescope camera aiming point, a calibration algorithm based on image processing has been implemented using a cross-correlation template matching operating in Fourier domain. This algorithm set the overlapping area of the high-magnified image on the large field of view. The shift between the two cameras alignment is caused by a margin of error while printing. Also, an image processing based tracking algorithm has been implemented using an adaptive background subtraction coupled to a Kalman filter. This tracking system aims to detect a celestial object, and move automatically the telescope is order to get the targeted object onto the overlapping area defined by the calibration system.
Subjects: Telescopes
Optical instruments
Night vision devices
Advisor(s): Romuald Jolivot
Waleed S. Mohammed
URI: http://dspace.bu.ac.th/jspui/handle/123456789/3912
Appears in Collections:Theses

Files in This Item:

File Description SizeFormat
Gael.pdf9.56 MBAdobe PDFView/Open
View Statistics

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.


  DSpace Software Copyright © 2002-2010  Duraspace - Feedback