ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

Development a Vision Based Seam Tracking System for None Destructive Testing Machines

Journal: International Journal of Computer Science and Network Solutions (IJCSNS) (Vol.1, No. 4)

Publication Date:

Authors : ;

Page : 45-56

Keywords : NDT; seam tracker; feature extraction; morphological filter; weld seam;

Source : Downloadexternal Find it from : Google Scholarexternal


The automatic weld seam tracking is an important challenge in None Destructive Testing (NDT) systems for welded pipe inspection. In this Study, a machine vision based seam tracker, is developed and implemented, instead of old electro-mechanical system. A novel algorithm based on the weld image centered is presented, to reduce Environment conditions and improve the seam tracking accuracy. The weld seam images are taken by a camera arranged ahead of the machine and the centered is extracted as a parameter to detect the weld position, and offset between this point and central axis is computed and used as control parameter of servomotors. Adaptive multi step segmentation t technique is employed to increase the probable of real edge of the welds and improve the line fitting accuracy. This new approach offers some important technical advantages over the existing solutions to weld seam detection: It’s based on natural light and does not need any auxiliary light. The adaptive threshold segmentation technique applied, decrease Environmental lighting condition. It's accurate and stable in real time NDT testing machines. After a series of experiments in real industrial environment, it is demonstrated that accuracy of this method can improve the quality of NDT machines. The average tracking error is 1.5 pixels approximately 0.25mm..

Last modified: 2014-01-30 15:32:28