Automated detection of welding defects in radiographic images
Date of Issue2007
School of Mechanical and Aerospace Engineering
An automatic detection system to detect welding defects in radiographic images is presented. Two new segmentation algorithms, MEWT and MTFEGA are proposed. They can succeed in segmenting welding defects presented in radiography, contrary to conventional methods. A classification system based on the support vector machine is proposed to the defect recognition.
DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
Nanyang Technological University