Vision-based automatic leak detection for short pipes in waterworks industry

Journal article

Authors/Editors

Strategic Research Themes

Publication Details

Author list: S. Vongbunyong; L. Floolek; P. Rodpai; P. Meena; K. Thamrongaphichartkul; A. Piyasinchart

Publication year: 2021

Journal: Journal of Research and Applications in Mechanical Engineering (2229-2152)

Volume number: 9

Issue number: 1

Start page: 1

End page: 8

Number of pages: 8

ISSN: 2229-2152

URL: https://ph01.tci-thaijo.org/index.php/jrame/article/view/243567

Abstract

One of the most common and widely used products in waterworks solution are short pipes. A number of manufacturing processes are involved in the production phase of these workpieces. Defection can be found in various forms in regard to the manufacturing processes, from the material forming to machining process. One of the most common defection found is related to the material porosity which leads to the leak problem. In this research, automatic leak detection system is designed and developed in order to inspect short pipes. The workpieces are submerged under water and bubbles will be detected by the vision-based detection algorithm to justify the leak. From the experiments, the inspection process took 15 seconds. The detection exhibited 0% false positive and 10% false negative. In addition, the production line has been transformed to a continuous line so that the the number of operators is reduced from 3 to 1.

Keywords

Bubbles inspection, Leak detection, Waterworks