Researches And Implementations Of Fast Ellipse Detector Based On Projective Invariant Pruning

The rapid ellipse detection method has a wide range of applications in the field of computer vision and multimedia.But with the increase of the amount of data processing,the computer vision application system has become more and more demanding of the efficiency of elliptical detection.At this time,how to detect oval objects quickly from the real image become a core problem.In the case of limited computing resources,and considering the complex relationships of multiple ellipses in real images and the presence of large numbers of candidate elements(edge points or arcs)for fitting ellipses in images,the existing method is difficult to apply to some real-time systems.In this dissertation,we present a fast algorithm detecting ellipses with high accuracy.The algorithm leverage a newly developed projective invariant to significantly prune the undesired candidates and to pick out elliptical ones.The invariant is able to reflect the intrinsic geometry of a planar curve,giving the value of-1 on any three collinear points and+1 for any six points on an ellipse.Thus,we apply the pruning and picking by simply comparing these binary values.Moreover,the calculation of the invariant only involves the determinant of a 3 × 3 matrix.Therefore,the algorithm can quickly select the desired candidate from a large number of candidate cases.Extensive experiments on three challenging data sets with 648 images demonstrate that our detector runs 20%-40% faster than the state-of-the-art algorithms with the comparable or higher precision.The threshold selection in the proposed method is verified on a complex real image database.This dissertation generates two synthetic databases,the robustness of the proposed method to elliptic variability is tested on 19100 synthetic image databases.