Star Identification Algorithm for Uncalibrated, Wide Fov Cameras

Abstract

A novel method is proposed for star identification via uncalibrated cameras with wide fields of view (FOVs). In this approach some of the triangles created by the stars in the FOV are selected for pattern recognition. The triangles are selected considering the sensitivity of their interior angles to the calibration error. The algorithm is based on the intersection between sets of triangles that are found in the database for each selected triangle of the image. By this method, most of the image stars contribute to pattern recognition and thereby it is very robust against the noise and the calibration error. The algorithm is performed on 150 night sky images, which are taken by an uncalibrated camera in FOV of 114°± 12°with a success rate of 94% and no false positives. Based on the identification approach, an adaptive method is also developed for calibrating and obtaining the projection function of an uncalibrated camera.

Non-Dimensional Star-Identification

“…The non-dimensional algorithm presented here is reminiscent of the algorithms presented in References [ 9 , 10 , 11 , 12 , 13 , 16 ]. Nevertheless, the non-dimensional algorithm introduced in this work has three major differences when compared with these previous algorithms: The NDSIA identifies stars using the dihedral angles of spherical star triangles.…”

“…Nevertheless, the non-dimensional algorithm introduced in this work has three major differences when compared with these previous algorithms: The NDSIA identifies stars using the dihedral angles of spherical star triangles. Previous algorithms used angles of planar star triangles [ 9 , 10 , 11 , 12 ], the sides of planar star triangles [ 13 ], or asterisms composed of four or five stars [ 16 ]. The NDSIA uses the n -dimensional k -vector (NDKV) [ 14 ] to search among a three-dimensional database.…”

“… The NDSIA uses the n -dimensional k -vector (NDKV) [ 14 ] to search among a three-dimensional database. Previous algorithms search among a one- or two-dimensional database [ 9 , 10 , 11 , 12 , 13 ], using searching techniques such as the one-dimensional k -vector [ 3 ], or by geometrically hashing asterisms of stars and using a hash-based search algorithm [ 16 ]. The NDSIA performs a final check that uses the identified stars’ interstellar angles to further the confidence in the Star-ID.…”

“…To that end, previous algorithms [ 9 , 10 , 11 , 12 , 13 ] have proposed solutions based on the non-dimensional Star-ID problem, i.e., the identification of stars when the focal length or OA offset of the star tracker were perturbed from their nominal values. However, these algorithms present an important limitation: the non-dimensional database used to obtain the Star-ID must be small.…”

Estimating camera parameters from starry night photographs

“…Estimation also takes a long time, often requiring tens of seconds. Recently, Ajdad et al proposed a star identification algorithm for images captured with wide field of view (FOV) cameras [21]. It takes into account errors caused by lens distortion of the wide FOV camera.…”