Design And Study Of Zipper Teeth-Numbers Real-Time Detection Based On Machine Vision

In the production process of zipper, because of production technology and so on there are always many deficiencies of proper numbers of teeth and this can result in degraded products of zipper. The factories of zipper usually sample in the process of producing or find out the degraded products of zipper after the producing, using handwork. This method has a lower efficiency, higher cost and the quality often affected by the by Human Factor. So it is not an ideal method.To solve this problem, a real-time zipper distribute detecting system based on PC is put forward in this paper. And then considering the special profiles of the zipper pictures, the proper picture processing arithmetic was chosen e.g. gray level transformation, histogram equalization, median filter, segmentation -thresholding and edge detect. An online real-time zipper distribute detecting system is proposed base on these arithmetic. The software of corresponding image processing has been developed base on the environment of VB. At last, the platform of real-time zipper distribute detecting system is established. And the detailed design proposal of communication interface between PC and SCM, the control interface between step motor and SCM and the beep alarm interface is given in this paper.Using this method we detected three class six types' zippers, analyze the real-time quality and the precision of this detection method, the result indicated that we can exactly detect the distribution of zipper fast. This has an important sense for developing the precise of detection of the process of producing.