A low complexity and high performance real-time algorithm of detecting and tracking circular shape in hole-punching machine
Wang Yi-gang, Liu Yangguang, Zhuoyuan Wang, Shengli Fan, Jialin Cui · 2008
In this paper, we proposed a real-time algorithm to detect and track the target circle based on region growing and accurately locating by multi-resolution. The proposed algorithm consists of edge detecting, region growing, the possible centre of circle estimating and the centre of circle accurate locating in turn. Firstly, we use Prewitt operator to form the edge of image and divide the formed edge into different connected sub-region by region growing method. Secondly, we compute the connected sub-regions’ centers of gravity, and take the centers as possible region of the target circle center. Finally we determine the accurate location by multi-resolution method. Experimental results show that all steps of the proposed algorithm consume less than 30ms while the location is accurate, and the proposed real-time algorithm can be used for hole-punching machine in practice.