Abstract The space environment is becoming more and more severe and crowded because of the rapid growth of space objects, which reveals an urgent demand to protect active satellites and other space assets. To accomplish such missions, e.g. the collision warning, the identification of space objects is important. In this paper, a three-stage approach for autonomous space object identification based on optical images is proposed. Firstly, on the basis of the approximate perspective imaging model, a scale and illumination invariant descriptor, composed of the normalized affine moment invariants (AMI) and the illumination invariant multiscale autoconvolution (MSA) transform, is developed to characterize the space object. Secondly, a multi-view modeling method is applied to construct multi-view databases of space objects for handling the viewpoint change. Finally, considering the extensibility of the databases, a K-nearest neighbor classifier is employed, and a K-means clustering is adopted to boost the search speed. Furthermore, to test the performance, a novel system based on the proposed approach is built and evaluated. The experimental evidence suggests that the system is stable and works well when the scale of a space object, the phase angle and the viewpoint change.

    Highlights ► Space objects are described using normalized AMI and illumination invariant MSA. ► Multi-view databases of space objects are constructed to handle viewpoint changes. ► KNN classifier identifies space objects, and K-means clustering boosts the speed. ► The method is effective when space object scale, phase angle and viewpoint change.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    An approach for autonomous space object identification based on normalized AMI and illumination invariant MSA


    Contributors:
    Ding, Hao (author) / Li, Xudong (author) / Zhao, Huijie (author)

    Published in:

    Acta Astronautica ; 84 ; 173-181


    Publication date :

    2012-11-22


    Size :

    9 pages




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    English





    Illumination-Invariant Color Object Recognition via Compressed Chromaticity Histograms of Color-Channel-Normalized Images

    Drew, M. S. / Wei, J. / Li, Z.-N. et al. | British Library Conference Proceedings | 1998


    A Chromaticity Space for Specularity-, Illumination Color- and Illumination Pose-Invariant 3-D Object Recognition

    Berwick, D. / Sang Wook Lee / IEEE; Computer Society | British Library Conference Proceedings | 1998


    Illumination Invariant Background Extraction

    Durucan, E. / Snoeckx, J. / Weilenmann, Y. et al. | British Library Conference Proceedings | 1999


    Enhanced illumination-invariant imaging

    HERMAN DAVID MICHAEL / VAIDYA AKSHAY / SANDERS LARRY | European Patent Office | 2024

    Free access