Subpixel edge extraction of part ant colony optimization-based and dimensional measurement

Lixiong Gong¹, Xiangsheng Kong², Yong Liu³, Min Huang⁴

COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(5) 240-246

¹School of Mechanical Engineering, Chongqing University of Technology, Chongqing 400054
²Department of Computer and Information Engineering, Xinxiang University, Xinxiang He’nan 453003
³School of Construction Machinery, Chang’an University, Xi’an Shanxi 710064
⁴MBA Education Center, Chongqing University of Technology, Chongqing 400050

Put forward a method combined improved ant colony and Zernike moment to detect image subpixel edge aiming at traditional ant colony algorithm’s drawback of long time consumption and easily to be affected by noise. The methods improved parameters from clustering centre setting, clustering operator selecting and pheromone updating, then extracted subpixel image edge based on Zerinke moments. Therefore, the result of image edge extraction is good and effective. Lastly, least square fitting is used to locate coordination of image edge and bearing of SKF 32308 J2/Q dimensions such as inner and outer diameters were measured. The result shows that the algorithm proposed can well to measure circular parts dimensions and has high precision.