裝置於衛星上的相機拍攝星空之影像，星體追蹤器利用這些影像來估算衛星之姿態及角速度。星體追蹤器擁有高彈性與高精度，其具有尺寸輕小、低計算成本、較廣運作範圍與軌道、可長時間運行，和角分精確度的特性。影像輸入後，首先通過影像處理處理雜訊，再計算出影像上每個似星點的中心位置作為特徵點，而後進入星體辨識的階段，利用上述特徵點與星體資料庫進行圖形匹配，最後使用匹配結果便能估算出衛星的姿態。星體辨識包含兩個模式：迷航模式和追蹤模式，兩者的差異在於是否擁有初始姿態，執行迷航模式的時機為當立方衛星最初進入軌道，或失去姿態資訊時；追蹤模式則是利用初始姿態進行星體配對。
本論文的目標是開發迷航模式之整體演算法，經由軟體在環模擬去分析以及驗證其效能，使其足以應用於立方衛星的星體追蹤器。本研究針對兩種迷航模式之星體辨識方法，分別是金字塔法和投票法。模擬之星空影像輸入演算法，輸出估算姿態，分析不同環境下的結果，並使用穩定度及計算成本來比較兩種星體辨識方法之效能。經過本研究之發展與驗證，在普通的環境下，金字塔法和投票法皆擁有99.5%的成功率，準確度皆在角分以上，及金字塔法較投票法少1.4%運行時間；在有雜點的環境，金字塔法展現較高的匹配穩定性，而耗時則較易受雜點影響。

A star tracker is used to estimate a satellite’s attitude and angular rate using images of star fields photographed with a satellite-based camera. The images input into the star tracker are first processed to distinguish stars from noise through the use of image processing. Then, the centroids are computed as the feature points using the weight-sum method. Next, star identification is conducted using the feature points to match the image patterns to those in the star catalog. Finally, the satellite’s attitude is estimated using the QUaternion ESTimator (QUEST) method. There are two modes in star identification: the Lost-in-Space (LIS) mode and tracking mode. The system uses LIS mode when it doesn’t have initial attitude information; otherwise it uses tracking mode.
The objective of this paper is developing LIS mode for a star tracker for CubeSats by analyzing and verifying the performance of two star identification methods, the pyramid method and the voting method, in LIS mode with a software-in-the-loop simulation. The entire LIS algorithm is developed; the simulated stellar images are input into the system, and estimated attitudes are obtained. The performance in different environments are analyzed based on the robustness and the computation cost. After development and verification in this research, in a normal situation, both the pyramid method and voting method were found to have 99.5%success rates, with an accuracy higher than arc-minutes, and the computation time was about 1.4% faster using the pyramid method. In a harsh environment that has false stars, the pyramid method showed higher matching robustness than the voting method, but the computation time made this process susceptible to false stars.

[1] CubeSat. Available online: http://www.cubesat.org (accessed on 1 July 2019).
[2] K. M. Huffman. Designing Star Trackers to Meet Micro-Satellite Requirements. Massachusetts Institute of Technology, 2006.
[3] B. S. Carlson. Comparison of modern CCD and CMOS image sensor technologies and systems for low resolution imaging. Proc. IEEE Sensors, vol. 1, pp. 171 -176, 2002.
[4] G. J. Zhang. Star Identification Methods, Techniques and Algorithms. Springer-Verlag Berlin Heidelberg, 2017.
[5] C. Liebe. Accuracy Performance of Star Trackers–A Tutorial. IEEE Transactions on Aerospace and Electronic Systems, vol. 38, no. 2, pp. 587-599, 2002.
[6] C. R. McBryde. A Star Tracker Design for CubeSats. In Aerospace Conference, 2012 IEEE, pages 1-14. IEEE, 2012.
[7] D. Mortari, M. A. Samaan, C. Bruccoleri and J. L. Junkins. The pyramid star identification technique. Navigation, 2004, 51, 171-183.
[8] E. J. Groth. A Pattern-Matching Algorithm for Two-Dimensional Coordinate Lists. Astronomical Journal, Vol. 91, 1986, pp. 1244-1248.
[9] D. Mortari. Search-Less Algorithm for Star Pattern Recognition, Journal of the Astronautical Sciences, Vol. 45, No. 2, April-June 1997, pp. 179-194.
[10] M. Kolomenkin, S. Pollak, I. Shimshoni, and M. Lindenbaum. Geometric voting algorithm for star trackers. IEEE Transactions on Aerospace and Electronic Systems, 56 vol. 44, no. 2, 2008.
[11] C. Padgett and K. Kreutz-Delgado. A grid algorithm for autonomous star identification. IEEE Trans. Aerospace Electron. Syst., vol. 33, pp. 202-213, 1997.
[12] C. Cole and J. Crassidis. Fast star-pattern recognition using planar triangles. J. Guid. Control. Dynam., pp. 1283-1286, 1994.
[13] J. Jiang, G. J. Zhang, X. Wei and X. Li. Rapid Star Tracking Algorithm for Star Sensor. IEEE Aerosp. Electron. Syst. Mag. 2009, 24, 23–33.
[14] J. Li, X. Wei and G. Zhang. An extended Kalman filter-based attitude tracking algorithm for star sensors. Sensors, vol. 17, no. 8, pp. 1921, 2017.
[15] J. A. Tappe. Development of Star Tracker System for Accurate Estimation of Spacecraft Attitude. Master Thesis, Naval Postgraduate School, Monterey, CA, 2009.
[16] G. M. Lerner. Three-Axis Attitude Determination. J. Wertz, Ed. D. Reidel Publishing Co.: D. Reidel Publishing Co., 1978.
[17] M. D. Shuster and S. D. Oh. Three-Axis Attitude Determination from Vector Observations. Journal of Guidance and Control, 70–77, 1981.
[18] G. Wang, F. Xing, M. Wei, T. Sun, and Z. You. Optimization method of star tracker orientation for sun-synchronous orbit based on space light distribution. Appl. Opt. 56, 4480–4490, 2017.
[19] M. B. Dillencourt, H. Samet, and M. Tamminen. A general approach to connected-component labeling for arbitrary image representations. Journal of the ACM (JACM), 39(2):253~280, 1992.
[20] J. Meeus. Astronomical Algorithms, 2nd ed. Willmann-Bell, 1988.
[21] E. Heide, M. Kruijff, S. Douma and D. Oude-Lansink. Development and Validation of a Fast and Reliable Star Sensor Algorithm with Reduced Data Base. IAF Melbourne, Tech. Rep. IAF-98.A.6.05, 1998.