Classification of Welding Defects Using Gray Level Histogram Techniques via Neural Network.

Authors

Wael Khalifa, Production Engineering and Mechanical Design Department, Faculty of Engineering, Mansoura University, 35516 Mansoura, EgyptFollow
Ossama Abouelatta, Production Engineering and Mechanical Design Department, Faculty of Engineering, Mansoura University, 35516 Mansoura, EgyptFollow
Elamir Gadelmawla, Production Engineering and Mechanical Design Department, Faculty of Engineering, Mansoura University, 35516 Mansoura, EgyptFollow
Ibrahim Elewa, Production Engineering and Mechanical Design Department, Faculty of Engineering, Mansoura University, 35516 Mansoura, Egypt

Corresponding Author

Khalifa, Wael

Subject Area

Production Engineering and Mechanical Design

Article Type

Original Study

Abstract

Technological development accompanied the need to get a high-quality welding. The important industries such as oil and auto industries and other important industries need to rely on reliable welding operations; collapse as a result of this welding may mean a great loss in lives and money. This paper aimed to produce an automatic system to detect, recognize and classify welding cases (defects and no defects) in radiography images was described depending upon image histogram technique. Two main steps to do that, In the first step, image processing techniques, including converting color images to gray scale, filtering image, and resizing were implemented to help in the image array of weld images and the detection of weld defects. The second step, a proposed program was build in-house depending upon Matlab to classify and recognize automatically six types of weld defects met in practice, it is Porosity – Undercut – Lac of fusion – Crack – Slag –Cavity, plus the non-defect type. It was clear from the results that it can rely on this method significantly, reaching rates as well as the appointment of defects and no defects to about 94.3%.

Keywords

Welding Defects; Neural network; Computer Vision; X; ray; Image histogram

Recommended Citation

Khalifa, Wael; Abouelatta, Ossama; Gadelmawla, Elamir; and Elewa, Ibrahim (2020) "Classification of Welding Defects Using Gray Level Histogram Techniques via Neural Network.," Mansoura Engineering Journal: Vol. 39 : Iss. 4 , Article 11.
Available at: https://doi.org/10.21608/bfemu.2020.102839