[1]徐奎奎,董兆伟,孙立辉,等.基于深度学习的 GH4169 合金组织表面缺陷检测[J].计算机技术与发展,2022,32(09):208-213.[doi:10. 3969 / j. issn. 1673-629X. 2022. 09. 032]
　XU Kui-kui,DONG Zhao-wei,SUN Li-hui,et al.Surface Defect Detection of GH4169 Alloy Structure Based on Deep Learning[J].,2022,32(09):208-213.[doi:10. 3969 / j. issn. 1673-629X. 2022. 09. 032]
点击复制

基于深度学习的 GH4169 合金组织表面缺陷检测()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年09期
页码:
208-213
栏目:
新型计算应用系统
出版日期:
2022-09-10

文章信息/Info

Title:
Surface Defect Detection of GH4169 Alloy Structure Based on Deep Learning
文章编号:
1673-629X(2022)09-0208-06
作者:
徐奎奎1 ; 董兆伟1 ; 孙立辉1 ; 董中奇2 ; 姜军强3
1. 河北经贸大学 信息技术学院,河北 石家庄 050061
2. 河北工业职业技术大学 材料工程系,河北 石家庄 050091
3. 西安科技大学 机械工程系,陕西 西安 710054
Author(s):
XU Kui-kui1 ; DONG Zhao-wei1 ; SUN Li-hui1 ; DONG Zhong-qi2 ; JIANG Jun-qiang3
1. School of Information Technology,Hebei University of Economics and Trade,Shijiazhuang 050061,China
2. Department of Material Engineering,Hebei Vocational University of Industry and Technology,Shijiazhuang 050091,China
3. Department of Mechanical Engineering,Xi’an University of Science and Technology,Xi’an 710054,China
关键词:
GH4169 合金; 缺陷检测; 注意力机制; RetinaNet; 特征金字塔网络
Keywords:
GH4169 alloy; defect detection; attention mechanism; RetinaNet; feature pyramid network
分类号:
TP391. 4
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 09. 032
摘要:
镍铁基高温合金 GH4169 合金由于其良好的综合性能被广泛应用在航空、石化、核能等行业,其冶炼工艺复杂,制备工艺路线较长,因此在铸造过程中会不可避免地产生大量组织缺陷,这些缺陷会对合金的性能造成重要的影响。 为了消除合金组织表面缺陷,需要研究合金组织表面缺陷的分布和成因以此提高合金的冶炼技艺。 但传统人工检测 GH4169合金组织表面缺陷效率低、精度差,很难用于检测大棒材。 因此,为了实现组织表面缺陷的自动检测,在 RetinaNet 网络结构的基础上提出了一种 CA-RetinaNet 网络结构用于 GH4169 合金组织表面缺陷检测,该方法主要增强了网络检测小缺陷的能力。 首先,在特征提取网络中使用了 CA-Resnet 结构,引入轻型注意力机制对感兴趣目标进行特征权重增强,提高了含有目标通道的权重;然后对 RetinaNet 网络中的特征金字塔网络进行了优化,重新构建了特征金字塔网络的底层结构,以获取更大的特征图检测小缺陷。 利用 CA-RetinaNet 网络模型在 GH4169 合金组织表面缺陷数据集上进行检测实验,取得了较高的准确率,相较于原始 RetinaNet 网络,mAP 值提升了 8. 6% ,极大地提升了网络的检测精度。
Abstract:
The nickel-iron-based superalloy GH4169 is widely used in aviation,petrochemical,nuclear energy and other industries due to its good comprehensive properties. The smelting process is complicated and the preparation process route is long,so a large number of structural defects will inevitably occur during the casting process,and these defects will have an important impact on the properties of the alloy. In order to eliminate the surface defects of the alloy structure,it is necessary to study the distribution and causes of the surface defects of the alloy structure to improve the alloy smelting technology. However, the traditional manual detection of GH4169 alloy structure surface defects has low efficiency and poor accuracy,and it is difficult to detect large bars. In order to realize the automaticdetection of tissue surface defects,based on the RetinaNet network structure,a CA-RetinaNet network structure is proposed for surface defect detection of GH4169 alloy tissue. This method mainly enhances the ability of the network to detect small defects. First,the CA-Resnet structure is used in the feature extraction network,and the light attention mechanism is introduced to enhance the feature weight ofthe target of interest, and increase the weight of the target channel. Then the feature pyramid network in the RetinaNet network is optimized,and the underlying structure of the feature pyramid network is constructed to obtain a larger feature map to detect small defects. Using the CA-RetinaNet network model to test the surface defect data set of the GH4169 alloy structure has achieved a high accuracy rate. Compared with the original RetinaNet network,the mAP value is increased by 8. 6% ,which greatly improves the detection accuracy of the network.

相似文献/References:

[1]张燕 刘春.基于圆心定位的瓶口三圆周快速缺陷检测算法[J].计算机技术与发展,2009,(06):243.
　ZHANG Yan,LIU Chun.A Rapid Defect Detecting Algorithm of Bottle＇s Rim Based on Central Location and Three- Circle Method[J].,2009,(09):243.
[2]舒坚 胡茂林.基于Markov随机场模型的纹理图像的缺陷检测[J].计算机技术与发展,2006,(05):65.
　SHU Jian,HU Mao-lin.Detection of Defect Based on Texture of Markov Random Field Model[J].,2006,(09):65.
[3]邓忠华 刘飞.印刷品缺陷检测的初步方案[J].计算机技术与发展,2006,(10):171.
　DENG Zhong-hua,LIU Fei.Preliminary Scheme for Defect Inspection of Printed Material[J].,2006,(09):171.
[4]李雪晨 汪仁煌 艾星芳.改进的弯曲度算法在阶梯修剪检测中的应用[J].计算机技术与发展,2012,(03):166.
　LI Xue-chen,WANG Ren-huang,AI Xing-fang.Use of Improved Tortuosity Algorithm in Step Form Disfigurement Test[J].,2012,(09):166.
[5]杨杰,卢盛林,赵晓芳.机器视觉在钢化玻璃缺陷检测中的应用研究[J].计算机技术与发展,2013,(03):211.
　YANG Jie,LU Sheng-lin,ZHAO Xiao-fang.Application and Research of Machine Vision in Tempered Glass Defect Inspection[J].,2013,(09):211.
[6]陈星[],徐迎晖[],肖青海[][]. QR码印刷品缺陷检测[J].计算机技术与发展,2015,25(10):191.
　CHEN Xing[],XU Ying-hui[],XIAO Qing-hai[][]. Defect Detection of Printed QR Code Image[J].,2015,25(09):191.
[7]杜晓辉,刘霖,张静,等.火花塞端面缺陷自动检测算法设计[J].计算机技术与发展,2019,29(02):172.[doi:10．3969/j．issn．1673－629X．2019．02．036]
　DU Xiaohui,LIU Lin,ZHANG Jing,et al.Optoelectronic Inspection of Defects for Surface of Spark Plug[J].,2019,29(09):172.[doi:10．3969/j．issn．1673－629X．2019．02．036]
[8]刘钱超,董超群,张垚.基于容器技术的软件测试优化研究[J].计算机技术与发展,2019,29(04):13.[doi:10. 3969 / j. issn. 1673-629X. 2019. 04. 003]
　LIU Qian-chao,DONG Chao-qun,ZHANG Yao.Research on Optimization of Software Testing Based on Container Technology[J].,2019,29(09):13.[doi:10. 3969 / j. issn. 1673-629X. 2019. 04. 003]
[9]张洪波,隋文涛,袁摇 林,等.基于深度自编码器网络的压盖缺陷检测[J].计算机技术与发展,2022,32(02):143.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 023]
　ZHANG Hong-bo,SUI Wen-tao,YUAN Lin,et al.Capping Defect Detection Based on Deep Autoencoder Network[J].,2022,32(09):143.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 023]
[10]丁龙飞,曾水玲.基于多重注意力机制的电机磁瓦表面缺陷检测[J].计算机技术与发展,2022,32(12):194.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 029]
　DING Long-fei,ZENG Shui-ling.Detection of Surface Defects of Motor Magnetic Tiles Based on Multiple Attention Mechanisms[J].,2022,32(09):194.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 029]

更新日期/Last Update: 2022-09-10