A robust circle detection algorithm based on top-down least-square fitting analysis

Abstract

In this paper, we propose a robust and efficient circle detector, which achieves accurate results with a controlled number of false detections and requires no parameter tuning. The proposed algorithm consists of three steps as follows. First, we propose a novel edge point chaining method to extract Canny edge segments (i.e., contiguous and sequential chains of Canny edge points). Second, we split each edge segment into several smooth sub-segments, and detect candidate circles within each obtained sub-segment based on top-down least-square fitting analysis. Third, we employ Desolneux et al.’s method to reject the false detections. Experimental results demonstrate that the proposed method is efficient and more robust than the state-of-the-art algorithm EDCircles.