|
In order to offer a potential for extending the life of satellites and reducing the launch and operating costs, satellite
servicing including conducting repairs, upgrading and refueling spacecraft on-orbit become much more frequently.
Future space operations can be more economically and reliably executed using machine vision systems, which can meet
real time and tracking reliability requirements for image tracking of space surveillance system.
Machine vision was applied to the research of relative pose for spacecrafts, the feature extraction algorithm was the basis
of relative pose. In this paper fractal geometry based edge extraction algorithm which can be used in determining and
tracking the relative pose of an observed satellite during proximity operations in machine vision system was presented.
The method gets the gray-level image distributed by fractal dimension used the Differential Box-Counting (DBC)
approach of the fractal theory to restrain the noise. After this, we detect the consecutive edge using Mathematical
Morphology.
The validity of the proposed method is examined by processing and analyzing images of space targets. The edge
extraction method not only extracts the outline of the target, but also keeps the inner details. Meanwhile, edge extraction
is only processed in moving area to reduce computation greatly. Simulation results compared edge detection using the
method which presented by us with other detection methods. The results indicate that the presented algorithm is a valid
method to solve the problems of relative pose for spacecrafts.
|
