Two-stage convolutional neural network for road crack detection and segmentation. (30th December 2021)

• Record Type: Journal Article
• Authors: Nguyen, Nhung Hong Thi; Perry, Stuart; Bone, Don; Le, Ha Thanh; Nguyen, Thuy Thi
• Is Part Of: Expert systems with applications. Volume 186(2021)
• Publication Date: 2021-12-30
• Subjects: Convolutional Neural Networks -- Deep learning -- Crack detection -- Crack segmentation -- Crack condition survey; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2021.115718
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: A new two-stage architecture based on Convolutional Neural Networks. Take advantages of detection at the sample level and segmentation at the pixel level. The two-stage model effective for noisy, low-resolution images, and imbalanced data. Experiments and results for several crack road datasets. A new dataset of challenging crack road will be made available to the researchers. Abstract: Automatic detection of road cracks is an important task to support road inspection for transport infrastructure. Various methods have been proposed for road crack detection and segmentation, however, there is no established method for handling real road images that are noisy and of low quality. In this paper, a new method utilising a two-stage convolutional neural network (CNN) is proposed for road crack detection and segmentation in images at the pixel level. Our novel contribution is a framework where the first stage serves to remove noise or artifacts and isolate the potential cracks to a small area, and the second stage is able to learn the context of cracks in the detected area. This is hence more effective than learning over the entire original noisy image. Extensive experiments on real datasets including public sources and our collected dataset have been conducted. The experimental results show that the two-stage CNN model outperformed existing approaches, especially for noisy, low-resolution images, and imbalanced datasets. Our approach achieves an F1-measure of over 0.91 on three datasets.