An investigation into the design of a satellite based stereo imaging sensor and the use of automatic image matching in the production of digital elevation models
Master Thesis
1993
Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
University of Cape Town
Department
License
Series
Abstract
Two problems are addressed in this dissertation. They are the design of a micro-satellite based stereo imaging sensor and the automatic matching of digital stereo images for automatic cartography applications. The two problems are related; they are both components of a stereo vision system. The research was initially motivated by the decision of the Department of Electrical and Electronic Engineering of the University of Stellenbosch to develop and build an experimental micro-satellite, SUNSAT. The proposed payload included a high resolution multi-spectral stereo imaging sensor. The second problem was motivated by the desire to use an automatic matching system to process the images produced by the sensor. The investigation, into the sensor design, was divided into two parts. The first part investigated the feasibility of the sensor and the second part dealt with the development of a design specification. The investigation, into automatic matching, dealt with the degree to which a set of requirements could be met. These requirements relate to the accuracy, reliability, generality, predictability and complexity of the matching system. The effect of scene characteristics was also investigated. The results showed that it is possible to build a micro-satellite based stereo imaging sensor. The recommended sensor design included three spectral bands, an 8 bit analogue-to-digital converter and a focal length of 535 mm. Furthermore, it was found that a sub-pixel accuracy matching requirement can be met and that a matching reliability of 89.6 can be achieved. Finally, it was found that the best matching results are obtained in areas of high image variance and low disparity variance.
Description
Bibliography: leaves 106-110.
Keywords
Reference:
Moon, P. 1993. An investigation into the design of a satellite based stereo imaging sensor and the use of automatic image matching in the production of digital elevation models. University of Cape Town.