An ellipse feature tracking method based on the Kalman filter

In this paper, a method of ellipse feature tracking based on the Kalman filter is proposed. In the application of the video analysis, this method can predict the results of the next frame, eliminate the false detection of the ellipse, and estimate ellipse parameters using prediction results and ellipse detection results. Through computer simulation experiments and semi physical simulation experiments, it is proved that the ellipse parameter accuracy is further improved compared with the single frame ellipse detection algorithm, and the method does not affect the ellipse detection in real time.