對於大量天文資料之分散式小行星軌跡探索

Distributed Asteroid Track Discovery for Large Astronomical Data

黃柏軒(Po-Hsuan Huang)

指導教授 ： 蔡孟峰

國立中央大學/資訊電機學院/資訊工程學系/碩士(2015年)

摘要

小行星(asteroid)遍布於整個太陽系當中，受到重力吸引而造成其移動的現象(moving object)。隨著硬體設備技術的進步，天文資料的獲取越來越快速。於大量的望遠鏡觀測資料當中尋找出小行星軌跡，將可以幫助天文學家了解更深層的天文意義，以及對於軌跡的路徑是否對地球造成危害作更進一步的預測。然而天文觀測資料的龐大，透過手動或人工的方式計算軌跡過於費時費力，且精確性較低。
為了要減少硬碟的寫入寫出，解決記憶體不足的問題，本論文將導入處理大量資料的方法，採用分散式系統，運用分散式的演算法，以具更系統性的方法尋找小行星軌跡。本論文將使用分散式的檔案系統與資料庫來做為儲存的設備，以良好可靠性與擴充性作為考量。同時應用分散式的Hough Transform演算法，搭配Open Source的雲端計算環境，更有效率地尋找小行星軌跡。
　　在實驗的部分，本論文將於Palomar Transient Factory(PTF)天文觀測資料中找出小行星軌跡。本論文的方法明顯的讓軌跡探索的效率有所改進，對於未來新獲取的觀測資料也能進行有效率的更新，並提供視覺化的介面來觀察小行星軌跡的移動。

關鍵字

大量資料 ； 雲端運算 ； 小行星軌跡 ； Hough Transform演算法

並列摘要

Asteroids are spread in the whole Solar System. They are attracted by gravity which makes them a moving object. As the hardware device improve, acquiring astronomical data becomes faster. Discovering asteroid track in the large observed data can benefit astronomical researchers in understanding some astronomical phenomena and can be examined for further prediction to whether it causes disasters to Earth. Due to large astronomical data, finding out asteroid track by manual is absolutely hard, inefficient and low accuracy.
In order to reduce disk I/O overhead and solve the problem of memory insufficient. This research proposes to adopt a distributed Hough Transform method with Big-Data data management and cloud computing technique for systematically searching asteroid tracks. This research utilizes distributed file system and database as the storage based on good reliability and scalability. It also uses an open source cloud computing environment for more efficiency in searching asteroid tracks.
　　In the experiments, this research discovers asteroid tracks in Palomar Transient Factory astronomical observed data. The results represent that our method can improve the efficiency apparently. New observed data can be also updated efficiently and provide the interface of visualization for observing the moving of asteroid.

參考文獻

[1] R. Duda and P. Hart, “Use of the Hough Transformation to Detect Lines and Curves in Pictures”, Communications of the ACM CACM Homepage archive Volume 15 Issue 1, Pages 11-15, Jan. 1972.

[5] C. Zhai, M. Shao, B. Nemati, T. Werne, H. Zhou, S. G. Turyshev, J. Sandhu, G. Hallinan, and L. K. Harding, “Detection of A Faint Fast-moving Near-earth Asteroid Using The Synthetic Tracking Technique”, The Astrophysical Journal, 792:60 (14pp), 2014 September 1.

[7] P. Gural, J. Larsen, and A. Gleason, “Matched filter processing for asteroid detection”, Astronomical Journal vol. 130 pp. 1951-1960 2005.

[9] Palomar Transient Factory, http://www.ptf.caltech.edu/

[11] D. Han and E. Stroulia, “A Three-Dimensional Data Model in HBase for Large Time-Series Dataset Analysis”, MESOCA, pp. 47-56, 2012.