Deep Convolution Neural Networks 이용하여 결함 검출을 위한 결함이 있는 철도선로표면 디지털영상 재 생성 = Regeneration of a defective Railroad Surface for defect detection with Deep Convolution Neural Networks

다국어 초록 (Multilingual Abstract)

This study was carried out to generate various images of railroad surfaces with random defects as training data to be better at the detection of defects. Defects on the surface of railroads are caused by various factors such as friction between track binding devices and adjacent tracks and can cause accidents such as broken rails [14], so railroad maintenance for defects is necessary. Therefore, various researches on defect detection and inspection using image processing or machine learning on railway surface images have been conducted to automate railroad inspection and to reduce railroad maintenance costs. In general, the performance of the image processing analysis method and machine learning technology is affected by the quantity and quality of data. For this reason, some researches require specific devices or vehicles to acquire images of the track surface at regular intervals to obtain a database of various railway surface images [15, 16]. On the contrary, in this study, in order to reduce and improve the operating cost of image acquisition, we constructed the 'Defective Railroad Surface Regeneration Model' by applying the methods presented in the related studies of the Generative Adversarial Network (GAN) [1]. Thus, we aimed to detect defects on railroad surface even without a dedicated database. This constructed model is designed to learn to generate the railroad surface combining the different railroad surface textures and the original surface, considering the ground truth of the railroad defects. The generated images of the railroad surface were used as training data in defect detection network [2], which is based on Fully Convolutional Network (FCN) [3]. To validate its performance, we clustered and divided the railroad data into three subsets, one subset as original railroad texture images and the remaining two subsets as another railroad surface texture images. In the first experiment, we used only original texture images for training sets in the defect detection model. And in the second experiment, we trained the generated images that were generated by combining the original images with a few railroad textures of the other images. Each defect detection model was evaluated in terms of 'intersection of union(IoU)' and F1-score measures with ground truths. As a result, the scores increased by about 10~15% when the generated images were used, compared to the case that only the original images were used. This proves that it is possible to detect defects by using the existing data and a few different texture images, even for the railroad surface images in which dedicated training database is not constructed.

참고문헌 (Reference)

• 1 최우용, "차세대 고속열차의 레일표면 결함 검출 시스템" 대한전기학회 66 (66): 870-876, 2017
• 2 Radford, Alec, "Unsupervised representation learning with deep convolutional generative adversarial networks"
• 3 International Union of Railways (UIC), "UIC 725:Treatment of Rail Defects"
• 4 International Union of Railways (UIC), "UIC 712: Rail defects"
• 5 R Andersson, Robin, "Surface defects in rails-Potential influence of operational parameters on squat initiation"
• 6 Yongzhi Min, "Real time detection system for rail surface defects based on machine vision" Springer Science and Business Media LLC 2018 (2018): 2018
• 7 김현호, "Railroad Surface Defect Segmentation Using a Modified Fully Convolutional Network" 한국인터넷정보학회 14 (14): 4763-4775, 2020
• 8 Tieleman, Tijmen, "Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude" 4 : 26-31, 2012
• 9 Mao, Xudong, "Least squares generative adversarial networks" 2017
• 10 Salimans, Tim, "Improved techniques for training gans. Advances in neural information processing systems"
• 11 Isola, Phillip, "Image-to-image translation with conditional adversarial networks" 2017
• 12 Azzam Sleit, "Image Clustering using Color, Texture and Shape Features" 한국인터넷정보학회 5 (5): 211-227, 2011
• 13 Goodfellow, Ian, "Generative adversarial nets. Advances in neural information processing systems"
• 14 Jonathan Long, "Fully Convolutional Networks for Semantic Segmentation" 2015
• 15 Schafer, Darwin H., "Effect of train length on railroad accidents and a quantitative analysis of factors affecting broken rails" University of Illinois at Urbana-Champaign 2008
• 16 Mirza, Mehdi, "Conditional generative adversarial nets"
• 17 Xie, Saining, "Aggregated residual transformations for deep neural networks" 2017
• 18 Kingma, Diederik P., "Adam: A method for stochastic optimization"
• 19 Jinrui Gan, "A Hierarchical Extractor-Based Visual Rail Surface Inspection System" 17 (17): 7935-7944, 2017