Matching CCD images to a stellar catalog using locality-sensitive hashing  

作　　者：Bo Liu Jia-Zong Yu Qing-Yu Peng 

机构地区：[1]College of Information Science and Technology, Jinan University [2]Sino-French Joint Laboratory for Astrometry, Dynamics and Space Science, Jinan University

出　　处：《Research in Astronomy and Astrophysics》2018年第2期105-114,共10页天文和天体物理学研究（英文版）

基　　金：supported by the National Natural Science Foundation of China(U1431227);Guangzhou Science and Technology Planning Project(201604010037)

摘　　要：The usage of a subset of observed stars in a CCD image to find their corresponding matched stars in a stellar catalog is an important issue in astronomical research. Subgraph isomorphic-based algorithms are the most widely used methods in star catalog matching. When more subgraph features are provided, the CCD images are recognized better. However, when the navigation feature database is large, the method requires more time to match the observing model. To solve this problem, this study investigates further and improves subgraph isomorphic matching algorithms. We present an algorithm based on a locality-sensitive hashing technique, which allocates quadrilateral models in the navigation feature database into different hash buckets and reduces the search range to the bucket in which the observed quadrilateral model is located. Experimental results indicate the effectivity of our method.The usage of a subset of observed stars in a CCD image to find their corresponding matched stars in a stellar catalog is an important issue in astronomical research. Subgraph isomorphic-based algorithms are the most widely used methods in star catalog matching. When more subgraph features are provided, the CCD images are recognized better. However, when the navigation feature database is large, the method requires more time to match the observing model. To solve this problem, this study investigates further and improves subgraph isomorphic matching algorithms. We present an algorithm based on a locality-sensitive hashing technique, which allocates quadrilateral models in the navigation feature database into different hash buckets and reduces the search range to the bucket in which the observed quadrilateral model is located. Experimental results indicate the effectivity of our method.

关 键 词：astronomical databases： miscellaneous -- methods： data analysis -- techniques： imageprocessing 

分 类 号：P152[天文地球—天文学]