Malashin Ivan, Tynchenko Vadim, Nelyub Vladimir, Borodulin Aleksei, Gantimurov Andrei, Krysko Nikolay V, Shchipakov Nikita A, Kozlov Denis M, Kusyy Andrey G, Martysyuk Dmitry, Galinovsky Andrey
Artificial Intelligence Technology Scientific and Education Center, Department of Welding, Diagnostics and Special Robotics, Bauman Moscow State Technical University, 105005 Moscow, Russia.
Scientific Department, Far Eastern Federal University, 690922 Vladivostok, Russia.
Sensors (Basel). 2024 May 31;24(11):3563. doi: 10.3390/s24113563.
The paper introduces a computer vision methodology for detecting pitting corrosion in gas pipelines. To achieve this, a dataset comprising 576,000 images of pipelines with and without pitting corrosion was curated. A custom-designed and optimized convolutional neural network (CNN) was employed for binary classification, distinguishing between corroded and non-corroded images. This CNN architecture, despite having relatively few parameters compared to existing CNN classifiers, achieved a notably high classification accuracy of 98.44%. The proposed CNN outperformed many contemporary classifiers in its efficacy. By leveraging deep learning, this approach effectively eliminates the need for manual inspection of pipelines for pitting corrosion, thus streamlining what was previously a time-consuming and cost-ineffective process.
本文介绍了一种用于检测燃气管道点蚀的计算机视觉方法。为此，精心整理了一个包含576,000张有和没有点蚀的管道图像的数据集。采用了定制设计和优化的卷积神经网络（CNN）进行二元分类，区分腐蚀图像和未腐蚀图像。尽管与现有的CNN分类器相比，该CNN架构的参数相对较少，但仍实现了高达98.44%的显著高分类准确率。所提出的CNN在有效性方面优于许多当代分类器。通过利用深度学习，这种方法有效地消除了对管道点蚀进行人工检查的需求，从而简化了以前既耗时又成本低效的过程。
