Convolution neural networks for pothole detection of critical road infrastructure. (April 2022)

• Record Type: Journal Article
• Authors: Pandey, Anup Kumar; Iqbal, Rahat; Maniak, Tomasz; Karyotis, Charalampos; Akuma, Stephen; Palade, Vasile
• Is Part Of: Computers & electrical engineering. Volume 99(2022)
• Publication Date: 2022-04
• Subjects: Pothole detection -- Convolution neural networks -- Crowdsource data -- Accelerometer data -- Highway maintenance; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2022.107725
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: A well developed and maintained highway infrastructure is essential for the economic and social prosperity of modern societies. Highway maintenance poses significant challenges pertaining to the ever-increasing ongoing traffic, insufficient budget allocations and lack of resources. Road potholes detection and timely repair is a major contributing factor to sustaining a safe and resilient critical road infrastructure. Current pothole detection methods require laborious manual inspection of roads and lack in terms of accuracy and inference speed. This paper proposes a novel application of Convolutional Neural Networks on accelerometer data for pothole detection. Data is collected using an iOS smartphone installed on the dashboard of a car, running a dedicated application. The experimental results show that the proposed CNN approach has a significant advantage over the existing solutions, with respect to accuracy and computational complexity in pothole detection.