Sample generation of semi-automatic pavement crack labelling and robustness in detection of pavement diseases

• Published in: Electronics letters Vol. 55; no. 23; pp. 1235 - 1238
• Main Authors: Jia, Guohui, Song, Weidong, Jia, Di, Zhu, Hong
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 14.11.2019
• Subjects: convolutional neural nets; convolutional neural networks; crack detection; crack features; crack images; ellipse fitting method; feature extraction; generalisation ability; Image and vision processing and display technology; image segmentation; learning (artificial intelligence); manual marking method; modified C–V model; multiscale feature extraction module; pavement diseases; rich deep convolutional features; road building; road cracks; sample generation; semiautomatic pavement crack labelling algorithm; structural engineering computing; generalisation ability; structural engineering computing; road cracks; pavement diseases; crack features; rich deep convolutional features; semiautomatic pavement crack labelling algorithm; road building; crack images; modified C–V model; ellipse fitting method; feature extraction; image segmentation; multiscale feature extraction module; sample generation; convolutional neural nets; manual marking method; convolutional neural networks; learning (artificial intelligence); crack detection
• ISSN: 0013-5194; 1350-911X; 1350-911X
• DOI: 10.1049/el.2019.2692

Recent convolutional neural networks have made significant advancements in the detection of road cracks. However, the lack of accurate crack training data reduces the generalisation ability of the deep model. In this Letter, a semi-automatic pavement crack labelling algorithm is proposed to solve the problem of insufficient training data. First, the modified C–V model is used to obtain the preliminary segmentation results. Second, the direction of the initial segmentation area is calculated by the ellipse fitting method, and the preliminary segmentation results are used as samples for accurate labelling. Finally, a multi-scale feature extraction module is proposed for learning rich deep convolutional features, which allows the acquired crack features under a complex background to be more discriminant. The experimental results were compared with the manual marking method, and this method can achieve accurate marking of crack images with a low amount of interaction, thereby significantly reducing the cost of ground-truth making. The results of the validation and comparison experiments on test data sets indicate that the proposed method can not only effectively identify cracks, but also overcome the interference of many factors in the environment.