TY - JOUR
T1 - Full field-of-view pavement stereo reconstruction under dynamic traffic conditions
T2 - incorporating height-adaptive vehicle detection and multi-view occlusion optimization
AU - Guan, Jinchao
AU - Yang, Xu
AU - Lee, Vincent C.S.
AU - Liu, Wenbo
AU - Li, Yi
AU - Ding, Ling
AU - Hui, Bing
N1 - Funding Information:
The study presented in the article was partially supported by the National Key Research and Development Program of China (No. 2021YFB2601000 ), National Natural Science Foundation of China (No. 52078049 ), Fundamental Research Funds for the Central Universities , CHD (No. 300102210302 , No. 300102210118 ) and the 111 Project of Sustainable Transportation for Urban Agglomeration in Western China (No. B20035 ). The authors would like to thank the Editor and the anonymous reviewers for their constructive comments and valuable suggestions on how to improve the quality of the article. The authors also appreciate the help of graduate students Kang Fan and Shuo Li for data acquisition and annotation.
Publisher Copyright:
© 2022 Elsevier B.V.
PY - 2022/12
Y1 - 2022/12
N2 - Three-dimensional (3D) surface information is becoming the data of choice for pavement inspection and maintenance. Pavement digital reconstruction still faces various challenges, including field-of-view (FOV) limits, traffic influences, data acquisition speed and cost. This paper presents a full-FOV pavement stereo reconstruction framework integrating unmanned aerial vehicle (UAV) photography, object detection and multi-view occlusion optimization. A UAV-YOLO vehicle detector embedding depth-wise separable convolution and resolution adjustment unit is developed for noise localization. To improve reconstruction quality and speed, multi-view occlusion optimization is proposed for determining the optimal image series spatial distribution. The results show that the UAV-YOLO detector achieves an overall AP75 of 93.69% with an inference speed of 157 FPS. Through multi-criterion evaluation, the pavement digital models reconstructed under dynamic traffic conditions have satisfactory performance in terms of point cloud noise, density, similarity and accuracy. In addition, the proposed stereo reconstruction workflow saves 30.27% processing time over conventional SfM workflow.
AB - Three-dimensional (3D) surface information is becoming the data of choice for pavement inspection and maintenance. Pavement digital reconstruction still faces various challenges, including field-of-view (FOV) limits, traffic influences, data acquisition speed and cost. This paper presents a full-FOV pavement stereo reconstruction framework integrating unmanned aerial vehicle (UAV) photography, object detection and multi-view occlusion optimization. A UAV-YOLO vehicle detector embedding depth-wise separable convolution and resolution adjustment unit is developed for noise localization. To improve reconstruction quality and speed, multi-view occlusion optimization is proposed for determining the optimal image series spatial distribution. The results show that the UAV-YOLO detector achieves an overall AP75 of 93.69% with an inference speed of 157 FPS. Through multi-criterion evaluation, the pavement digital models reconstructed under dynamic traffic conditions have satisfactory performance in terms of point cloud noise, density, similarity and accuracy. In addition, the proposed stereo reconstruction workflow saves 30.27% processing time over conventional SfM workflow.
KW - Dynamic traffic influence
KW - Full field of view
KW - Multi-view optimization
KW - Object detection
KW - Pavement digital model
KW - Stereo vision
KW - Structure from motion (SfM)
KW - UAV photography
UR - http://www.scopus.com/inward/record.url?scp=85140138105&partnerID=8YFLogxK
U2 - 10.1016/j.autcon.2022.104615
DO - 10.1016/j.autcon.2022.104615
M3 - Article
AN - SCOPUS:85140138105
SN - 0926-5805
VL - 144
JO - Automation in Construction
JF - Automation in Construction
M1 - 104615
ER -