Quality-of-service metrics for evaluating sensor fusion systems without ground truth

This paper presents some results of sensor fusion in maritime domain, applying a track fusion solution to real datasets containing local tracks from multiple radar and AIS. Some performance metrics are proposed to assess the quality of fusion output. The capability of track fusion logic to solve some typical problems of local trackers is shown and how the contextual knowledge of radars occlusions areas can be refined to improve the global performance.

[1]  Paolo Braca,et al.  Maritime surveillance with multiple over-the-horizon HFSW radars: An overview of recent experimentation , 2015, IEEE Aerospace and Electronic Systems Magazine.

[2]  Jesús García,et al.  Robust sensor fusion in real maritime surveillance scenarios , 2010, 2010 13th International Conference on Information Fusion.

[3]  Erik P. Blasch,et al.  Fusion metrics for dynamic situation analysis , 2004, SPIE Defense + Commercial Sensing.

[4]  Fakhri Karray,et al.  Multisensor data fusion: A review of the state-of-the-art , 2013, Inf. Fusion.

[5]  James Llinas,et al.  An introduction to multisensor data fusion , 1997, Proc. IEEE.

[6]  Gaston H. Gonnet,et al.  Handbook of Algorithms , 1984 .

[7]  Moshe Idan,et al.  Advances in Estimation, Navigation, and Spacecraft Control , 2015 .

[8]  Hervé Fargetton,et al.  Control of multi sensor system based on anomaly mapping and expert system , 2011, GI-Jahrestagung.

[9]  Jesús García,et al.  Geographic context configuration in fusion algorithms for maritime surveillance , 2014, 17th International Conference on Information Fusion (FUSION).