摘要: |
为了检测X射线探伤图像中螺旋钢管焊缝缺陷,提出了一种新的实时自动检测算法。它是一种基于图像空间特性(空间对比度与空间方差)的模糊识别算法,与人眼视觉识别特性相近,与其它传统缺陷检测算法相比,有较低的漏判率和误判率,并能满足生产线上实时检测的要求。 |
关键词: 无损探伤 焊缝缺陷 自动检测 模糊算法 |
DOI: |
分类号: |
基金项目: |
|
Real-time automatic detection of weld defects in X-ray images |
SUN Yi1, SUN Hong-yu1, BAI Peng1, WANG Yu2, TIAN Yan-ping2
|
1.School of Electronic and Information Engineering, Dalian Univ of Tech, Dalian 116024, China;2.Steel Tube Factory of China Oil & Gas Pipeline Burean, Liaoning Liaoyang 111000, China
|
Abstract: |
A new real-time automatic detection algorithm is taken up for applying to X-ray images of screw steel tube. As a fuzzy method based on spatial contrast and differece in images, it can achieve lower missing rate and fewer misinterpretations than other inspecting algorithm, and it can satisfy the requirements of real-time detection. |
Key words: NDT weld defect automatic inspection fuzzy algorithm |