A Fast and High-Precision Ellipse Detection Method Based on the Candy's Theorem

Ellipse detection is one of the important parts of intelligent manufacturing, and the efficient detection of elliptical contours corresponding to hole parts is the prerequisite before performing practical applications. Aiming at enhancing the efficiency in ellipse detection, a fast and high-precision ellipse detection method based on the Candy's theorem is proposed, in which the constrained simple polygon constraint and the Candy's theorem constraint are used to improve the efficiency of ellipse detection. The theory of the proposed ellipse detection method is described in detail. The comparison experiments with six popular ellipse detection methods are performed based on three public datasets, which are to verify the ellipse detection ability of the proposed method in improving detection speed and precision. And a small hole detection experiment is performed with the constructed experimental setup, which is to verify the effectiveness of the proposed method to detect small holes in practical applications. The proposed method can provide referenced methods and theories in the field of ellipse detection.