Autonomous detection of damage to multiple steel surfaces from 360° panoramas using deep neural networks

Luo, Cai and Yu, Leijian and Yan, Jiaxing and Li, Zongwei and Ren, Peng and Bai, Xiao and Yang, Erfu and Liu, Yonghong (2021) Autonomous detection of damage to multiple steel surfaces from 360° panoramas using deep neural networks. Computer-Aided Civil and Infrastructure Engineering, 36 (12). pp. 1585-1599. ISSN 1467-8667 (https://doi.org/10.1111/mice.12686)

[thumbnail of Luo-etal-CACIE-2021-Autonomous-detection-of-damage-to-multiple-steel-surfaces-from-360-panoramas]
Preview
Text. Filename: Luo_etal_CACIE_2021_Autonomous_detection_of_damage_to_multiple_steel_surfaces_from_360_panoramas_.pdf
Accepted Author Manuscript

Download (21MB)| Preview

Abstract

Structural health assessments are essential for infrastructure. By using an autonomous panorama vision‐based inspection system, the limitations of the human cost and safety factors of previously time‐consuming tasks have been overcome. The main damage detection challenges to panorama images are (1) the lack of annotated panorama defect image data, (2) detection in high‐resolution images, and (3) the inherent distortion disturbance for panorama images. In this paper, a new PAnoramic surface damage DEtection Network (PADENet) is presented to solve the challenges by (a) using an unmanned aerial vehicle to capture panoramic images and a distorted panoramic augmentation method to expand the panoramic dataset, (b) employing the proposed multiple projection methods to process high‐resolution images, and (c) modifying the faster region‐based convolutional neural network and training via transfer learning on VGG‐16, which improves the precision for detecting multiple types of damage in distortion. The results show that the proposed method is optimal for surface damage detection.

ORCID iDs

Luo, Cai, Yu, Leijian, Yan, Jiaxing, Li, Zongwei, Ren, Peng, Bai, Xiao, Yang, Erfu ORCID logoORCID: https://orcid.org/0000-0003-1813-5950 and Liu, Yonghong;