A New Star Identification Algorithm Based on Fuzzy Algorithms

Abstract

AbstractThe proposed algorithm is for satellite attitude determination. In this algorithm, the star point pattern convert to line pattern by “Delaunay triangulation” method, then we present a fuzzy line pattern matching. In this method, we use the membership functions to describe position, orientation and relation similarities between different line segments. The simulation results based on the “Desktop Universes Star images” demonstrate that the fuzzy star pattern recognition algorithm speeds up the process of star identification and increases the rate of success greatly (96.4%) compared with traditional matching algorithms. In addition, since the quality of star images play an important role in improving accuracy of star pattern recognition algorithm, therefore for image pre-processing we propose a fuzzy edge detection technique. This method highly affects noise cancellation, star features extraction, database production and matching algorithm.KeywordsMembership FunctionDelaunay TriangulationLine PatternAttitude DeterminationNoise CancellationThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.