摘要
提出了一种管路焊缝自动检测算法,通过计算管路图像的梯度-方向直方图确定管路的主方向,解决了传统的旋转穷举法计算量巨大的问题;通过HSV颜色空间转换增强了焊缝的特征,提高了模板匹配的鲁棒性和定位精度。实验表明该算法能够实现快速的焊缝自动检测和准确定位。
An automatic weld seam detection algorithm is proposed in this paper. The main direction of the pipeline is determined by the gradient-direction histogram, which saves huge amount of computation comparing to the traditional algorithm. The saturation in the HSV color space is adopted to enhance the feature of weld seam which increases robust of detection and the precision of the weld seam localization. A series of experiments show that our algorithm achieves high accuracy in the weld seam detection and localization.
出处
《航天器环境工程》
2014年第4期409-414,共6页
Spacecraft Environment Engineering
基金
北京卫星环境工程研究所自主研发课题(项目编号:YY-BZ(WM)-2013-18)
关键词
焊缝识别
HSV颜色空间
梯度-方向直方图
自动检测
weld seam detection
HSV color space
gradient-direction histogram
automatic detection
