摘要
为了提高石油钻机焊缝识别的精度,提出一种基于改进Mask基于区域的卷积神经网络(RCNN)的识别方法。以Mask RCNN为基础,改进Mask RCNN的特征网络,并引入卷积块注意力模块(CBAM)注意力机制进行权重灵活分配,构建改进Mask RCNN的石油钻机焊缝识别结构,实现对石油钻机焊缝的识别。结果表明,所提出的方法可实现不同类型的石油钻机焊缝识别,且识别准确率、精确率、交并比分别为91.94%、92.44%、91.87%。由此得出,所提出的方法可提高对石油钻机焊缝的识别精度,实现钻机焊缝的无损检测。
To improve the weld seam recognition precision of oil drilling rigs,a recognition method based on improved Mask region-based convolutional neural network(RCNN)is proposed.Based on Mask RCNN,the feature network of Mask RCNN is improved,convolutional block attention module(CBAM)attention mechanism is introduced for flexible weight allocation,and weld seam recognition structure of oil drilling rigs based on improved Mask RCNN is constructed to realize the recognition of weld seam of oil drilling rigs.The results show that the proposed method can achieve the recognition of different types of weld seam of oil drilling rigs,and its recognition accuracy,precision and intersection over union are 91.94%,92.44%and 91.87%,respectively.It is concluded that the proposed method can improve the recognition precision of weld seam of oil drilling rigs and realize the non-destructive testing of drilling rigs.
作者
沈学峰
徐晓磊
SHEN Xuefeng;XU Xiaolei(Liupu Drilling Branch of Sinopec East China Petroleum Engineering Co.,Ltd.,Zhenjiang 212003,China;RG Petro-machinery Group Co.,Ltd.,Nanyang 473006,China)
出处
《微型电脑应用》
2025年第12期200-203,共4页
Microcomputer Applications
关键词
无损检测
石油钻机
焊缝识别
Mask
RCNN
注意力机制
non-destructive testing
oil drilling rigs
weld seam recognition
Mask RCNN
attention mechanism
