Fast Ellipse Detection and Automatic Marking in Planar Target Image Sequences

Abstract

A fast method of detecting and automatic marking ellipses in the circle array target image sequences with high quality and strong regular pattern is proposed in this paper. This method collects a small amount of edge points from the ellipse by directly scanning the target image and searching for jumpy-changing of pixels, which can be used to identify the rest of other edge points by combining with the block properties of the elliptical regions, so that the expected ellipses can be calculated. The ellipse raster conversion process is then used to optimize and obtain the correct parameters of ellipses. After this, the multiple coordinate systems can be created based on the corner ellipses on the boundary of the ellipse array, and the fast marking of topological position of ellipses can also be achieved by mapping the ellipses between the different coordinate systems. The main advantage of the proposed method is that it significantly reduces the time complexity as it does not require the complex edge detection and spatial transformation operation. In addition, it is not restricted to any specific target models, and so it has very strong universality. This method is easy to be implemented, and has the high accuracy of detection. Thus, it can meet the requirements of the fast detection of the target image.