Meteoroid Recognition Algorithm of Civil Astronomical Meteor Monitoring System Based on Meteor Trajectory Recognition and Light Curve

Due to the decentralized nature of civil meteor monitoring and the lack of conditions for large-scale observations like in scientific astronomical meteor monitoring, the identification algorithms for scientific astronomical meteor monitoring cannot be applied to civil meteor monitoring. In response to the problem that the civil meteor monitoring system cannot effectively distinguish meteors from other moving objects during meteor recognition, a meteor candidate recognition algorithm has been designed and implemented. This algorithm mainly includes two parts: meteor trajectory recognition and light change curve shape analysis. By using this meteor recognition algorithm to process two months’ worth of images from the civil meteor monitoring system, 7400 meteor trajectories were extracted, of which 25% were non-meteor objects. The algorithm classifies the observed moving objects into target sets, and by considering the features of the light variation curve’s single-peak structure and multi-frame features, the final algorithm feature accuracy is 86%.