Estimation of pavement crack ratio by top-view transformation of in-vehicle smartphone camera and deep learning-based classification

• Published in: Artificial Intelligence and Data Science Vol. 3; no. 3; pp. 26 - 39
• Main Authors: GEDA, Jose Maria Guyamin, XUE, Kai, NAGAYAMA, Tomonori, ZHAO, Boyu, NAKA, Michihiro
• Format: Journal Article
• Language: English
• Published: Japan Society of Civil Engineers 2022
• Subjects: crack ratio; deep learning; image stitching; road condition assessment; top-view transformation

The conventional road crack ratio evaluation in Japan uses top-view images of pavement obtained from precision line scan cameras and manually classifies 0.5m grids of pavement surface based on the number of cracks. While inexpensive crack evaluation using in-vehicle smartphone cameras has been propo...