E-Zone: A Faster Neighbor Point Query Algorithm for Matching Spacial Objects

Abstract

Latest astronomy projects observe the spacial objects with astronomical cameras generating images continuously. To identify transient objects, the position of these objects on the images need to be compared against a reference table on the same portion of the sky, which is a complex search task called cross match. We designed Euclidean-Zone (E-Zone), a method for faster neighbor point queries which allows efficient cross match between spatial catalogs. In this paper, we implemented E-Zone algorithm utilizing euclidean distance between celestial objects with pixel coordinates to avoid the complex mathematical functions in equatorial coordinate system. Meanwhile, we surveyed on the parameters of our model and other system factors to find optimal configures of this algorithm. In addition to the sequential algorithm, we modified the serial program and implemented an OpenMP parallelized version. For serial version, the results of our algorithm achieved a speedup of 2.07 times over using equatorial coordinate system. Also, we achieved 19 ms for sequencial queries and 5 ms for parallel queries for 200,000 objects on a single CPU processor over a 230,520 synthetic reference database.