摘要
基于船体焊缝人工检测方式劳动强度大、效率低等现状,针对未融合、裂纹、气孔和焊瘤等典型焊缝缺陷,研究图像预处理和轮廓提取算法,对焊缝特征参数计算方法进行改进以降低图像处理运算量,提出基于有限样本的船体焊缝缺陷识别流程。最后应用OpenCV和C++编程,以焊缝缺陷正确识别率为标准,对图像处理算法和识别流程的有效性进行验证。
Based on the current situation of high labor intensity and low efficiency in the manual detection of hull welds,the image preprocessing and contour extraction algorithms are studied for typical weld defects such as nonfusion,cracks,blowholes,and solder bumps.The calculation method of image feature parameters is improved to reduce the amount of calculation,and a hull weld defect recognition process based on limited samples is proposed.Finally,OpenCV and C++ programming are used to verify the effectiveness of the image processing algorithm and recognition process based on the correct recognition rate of weld defects.
作者
何滨昂
刘玉良
HE Binang;LIU Yuliang(School of Information Science and Engineering,Hunan University,Changsha 410012,China;School of Information and Engineering,Zhejiang Ocean University,Zhoushan 316022,Zhejiang,China;Zhoushan Shangwen Robot Technology Co.,Ltd.,Zhoushan 316100,Zhejiang,China)
出处
《机电设备》
2021年第3期85-89,共5页
Mechanical and Electrical Equipment
基金
浙江省公益性项目(2015C31072)
舟山市科技局计划项目(2020C21001
2019C33103)。
关键词
船体焊缝
缺陷检测
图像处理
编程
hull weld
flaw inspection
image processing
programming