Title: Intelligent detection method for tapping omitting of internal thread based on computer vision

Authors: Wei Ding; Qingguo Wang; Yanfang Zhao

Addresses: School of Electrical Engineering and Automation, Changshu Institute of Technology, Changshu, 215500, China; Institute for Intelligent Systems, Faculty of Engineering and the Built Environment, University of Johannesburg, Johannesburg, South Africa ' Institute for Intelligent Systems, Faculty of Engineering and the Built Environment, University of Johannesburg, Johannesburg, South Africa ' School of Mechanical Engineering, Jiangsu University of Technology, Changzhou, 213000, China

Abstract: Tapping plays an important role in machining internal thread, whereas tapping omitting is inevitable in automated mass production. In this paper, a method for interior thread detection is presented for its efficiency, stability and robustness. Some key technologies such as imaging scheme and image processing algorithms were studied on the basis of the self-designed online detecting device. Firstly, the global threshold segmentation algorithm and local region growing segmentation algorithm are combined to precisely segment the imaging region of screw holes, while the imaging region of thread region is acquired based on the previous orientation. Then, the shape descriptors of the hole and texture descriptors of thread region are computed, respectively. And each descriptor's attribute importance is obtained using rough set. Finally, based on significant classification descriptors, semantic recognition of tapping omitting is achieved by heuristic rules. The experiment shows that the proposed classification algorithm using semantic features can identify the internal thread. The average detection time is 0.256 s, the detection accuracy of tapping omitting is 95.88%, and the detection accuracy of tapping thread is 100%.

Keywords: computer vision; tapping omitting; interior thread; semantic recognition.

DOI: 10.1504/IJMIC.2019.103657

International Journal of Modelling, Identification and Control, 2019 Vol.32 No.3/4, pp.238 - 243

Received: 11 Feb 2019
Accepted: 10 Mar 2019
Published online: 18 Nov 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article