Abbas Abaei Kashan - Department of Mechanical Engineering, Iran University of Science and Technology, Tehran, Iran;
Seyyed Mehrdad Tabatabaie - Maintenance Department, Civil Aviation Technology College, Tehran, Iran;
Mohammad Sadegh Alinia - Department of Metallurgical and Materials Engineering, Amirkabir University of Technology, Tehran, Iran;
Amid Maghsoudi - Department of Mechanical Engineering, Iran University of Science and Technology, Tehran, Iran;

Robots come to contribute human-being to make easy the most overwhelming works. They are able to fulfill mankind tasks, depending on programming precision, without any exhaustion. The crack finding discussion is likewise one of the most significant matters in Non-destructive testing science. According to this requirement, it is needed abundance of accuracy to be precise for having these tests done. In the current paper, design and construction of a robot which is able to recognize such cracks in welding spots has been represented. The stated robot are capable of withholding the located spot coordinates, taking photos of that welding place and recording them in its memory. Afterward, the images can be processed with getting helped by deep learning algorithm and in case of detecting crack, the presence and location as well would be asserted.

Non-destructive test robot, image processing, crack detection, welding spot.