Hybrid Grid Pattern Star Identification Algorithm Based on Multi-Calibration Star Verification

Area of Science:

• Astronomy
• Astrophysics
• Space Science

Background:

• Star identification is crucial for autonomous navigation in space.
• Existing algorithms struggle with small field-of-view cameras and noise.

Purpose of the Study:

• To develop a robust star identification algorithm for lost-in-space scenarios.
• To improve accuracy and reliability in challenging imaging conditions.

Main Methods:

• A hybrid grid pattern using multi-calibration stars for initial matching.
• Calibration star filtering to eliminate false positives.
• Nearest principle verification using reference star images.

Main Results:

• Achieved 96.43% identification rate with 2-pixel position noise.
• Achieved 96.45% identification rate with 0.3-magnitude noise.
• Successfully identified stars down to magnitude 12 in real tests.

Conclusions:

• The proposed hybrid grid algorithm outperforms existing grid-based methods.
• The algorithm offers improved robustness against position and magnitude noise.
• It is effective for scientific cameras in lost-in-space applications.