Abstract
This paper focuses on measuring weld pool geometry by removing interferences overlapped with the weld pool region in images. The images come from a vision system observing deep penetrate CO2 laser welding. The measurement is achieved through computer vision. First, Otsu’s method is used for threshold processing to get the binary image. Second, morphological opening with a specific structuring element is used to remove interferences and recover the weld pool. This method is called specific structuring element-based opening (SSEO). It effectively eliminates interferences like spatter, arc light, and plasma to obtain the accurate weld contour. The accurate contour is then used to calculate precise geometry parameters. The SSEO method is also efficient enough to satisfy the real-time requirements.
Original language | English (US) |
---|---|
Pages (from-to) | 1465-1477 |
Number of pages | 13 |
Journal | International Journal of Advanced Manufacturing Technology |
Volume | 90 |
Issue number | 5-8 |
DOIs | |
State | Published - May 1 2017 |
Keywords
- Computer vision
- Dilation and erosion
- Geometry parameter
- Image morphology
- Laser welding
- Weld pool
ASJC Scopus subject areas
- Control and Systems Engineering
- Software
- Mechanical Engineering
- Computer Science Applications
- Industrial and Manufacturing Engineering