維持・修繕（舗装）(2)

[V-514] Verification of applicability of pavement crack analysis technology to maintenance using general-purpose video camera images

〇Shigehisa Akiyama¹, Kyoichi Takeuchi¹, Ryuichi Imai², Kenji Nakamura³, Yoshinori Tsukada⁴, Nobuyuki Kumakura⁵, Yasuhiko Yamazaki⁶, Yoko Yonekawa⁶ (1.Nippon Koei Co., Ltd., 2.Hosei University, 3.Osaka University of Economics, 4.Setsunan University, 5.TOSHIBA ENVIRONMENTAL SOLUTIONS CORPORATION, 6.Toshiba Infrastructure Systems & Solutions Corporation)

Keywords:Pavement inspection technology, Crack rate, Soundness, Damage assessment, Deep learning

Visual inspection of pavement has many operational issues, such as qualitative evaluation and time-consuming preparation of records. In recent years, technology for detecting cracks by applying deep learning to a road surface image captured by a video camera has been developed. Using a video camera, the cracks on the national road were evaluated. The applicability of the technology to road management was evaluated by comparing the results of crack evaluation and the results of road surface property survey.