Crimson Publishers Publish With Us Reprints e-Books Video articles

Abstract

Research in Medical & Engineering Sciences

Ultra-Fast Detection of Loosened Bolts Based on Binary Image Processing

  • Open or CloseMir Hamid Rezaei1* and Javad Ostadieh2

    1Department of Electrical Engineering, Technical and Vocational University, Iran

    2Department of Communications and Electronics, Urmia University, Iran

    *Corresponding author: Mir Hamid Rezaei, Department of Electrical Engineering, Faculty of Ghazi Tabatabai, West Azarbaijan Branch, Technical and Vocational University, Urmia, Iran

Submission: May 05, 2021;Published: July 29, 2021

DOI: 10.31031/RMES.2021.09.000713

ISSN: 2576-8816
Volume 9 Issue 2

Abstract

In this paper, we offer a new method for detecting tight or loosened bolts based on image processing. The proposed method is independent of bolt appearances, and no height detector is needed. Firstly, several pictures are taken from the bolts in tight and different loosened states for training the system. After the training step, the status of the study bolts is processed by MATLAB software. By using image processing techniques including noise removal, Circular Hough Transform (CHT), transformation of colored images to grayscale, Otsu thresholding, binary complement of Otsu thresholding, enhancement of images with dilation structure element, and after calculating the area of the bolts, the decision about the states of the bolts can be accurately determined based on the number of zero pixels. The proposed method has the advantages of the low calculation process, very fast, and high precision response.

Keywords: Circle hough transform; Image processing; Inspecting; Loosened bolt; Otsu thresholding

Get access to the full text of this article