Article ID: | iaor201530575 |
Volume: | 110 |
Start Page Number: | 9 |
End Page Number: | 17 |
Publication Date: | May 2015 |
Journal: | Acta Astronautica |
Authors: | Sun Rong-yu, Zhan Jin-wei, Zhao Chang-yin, Zhang Xiao-xiang |
Keywords: | artificial intelligence, datamining |
The GEO region is a kind of unique and precious resource, it is of great importance to prevent space debris collisions in this region. Surveying space debris in GEO and extracting the relevant information automatically are rather challenging tasks due to several factors. Here we present an image processing pipeline which detects GEO objects automatically and improves the detection ability for faint objects. In our pipeline the mathematical morphology operator is adopted to eliminate noises and perform image restoration, then the median is used to eliminate the influences of field stars and extract object positions. The pipeline is tested on a large number of raw CCD images, and the tracklets obtained are correlated and catalogued, finally the effectiveness and the efficiency are demonstrated and proved by the results.