Center detection algorithm for printed circuit board circular marks based on image space and parameter space

A highly efficient circle positioning algorithm, called the two-step optimization Hough transform (TSHT), based on multi-resolution segmentation is proposed to solve problems of offset transform, namely, its large memory overhead, long time consumption, and low recognition accuracy. First, using image feature printed circuit board (PCB) circular identifier, target obtained adaptive preprocessing, then, images an acceptable quality are separated by shape inspection improve their robustness. Second, effective interval sampling strategies gradually controlling accumulative parameters, TSHT algorithm reduces overhead quickly locates center at pixel level. Finally, sub-pixel level found least-squares method for fitting. The experiments prove that TSHT, as a result high robustness, strong anti-noise capability, fast speed, accuracy, can be successfully applied vision system solder paste printing machine.