RISS 검색 - 국내학술지논문 상세보기

다국어 초록 (Multilingual Abstract)

Despite recent breakthroughs in deep learning and computer vision fields, the pixel-wise identification of tiny objects in high-resolution images with complex disturbances remains challenging. This study proposes a modified U-net for tiny crack segmentation in real-world steel-box-girder bridges. The modified U-net adopts the common U-net framework and a novel Self-Attention-Self-Adaption (SASA) neuron as the fundamental computing element. The Self-Attention module applies softmax and gate operations to obtain the attention vector. It enables the neuron to focus on the most significant receptive fields when processing large-scale feature maps. The Self-Adaption module consists of a multiplayer perceptron subnet and achieves deeper feature extraction inside a single neuron. For data augmentation, a grid-based crack random elastic deformation (CRED) algorithm is designed to enrich the diversities and irregular shapes of distributed cracks. Grid-based uniform control nodes are first set on both input images and binary labels, random offsets are then employed on these control nodes, and bilinear interpolation is performed for the rest pixels. The proposed SASA neuron and CRED algorithm are simultaneously deployed to train the modified U-net. 200 raw images with a high resolution of 4928 × 3264 are collected, 160 for training and the rest 40 for the test. 512 × 512 patches are generated from the original images by a sliding window with an overlap of 256 as inputs. Results show that the average IoU between the recognized and ground-truth cracks reaches 0.409, which is 29.8% higher than the regular U-net. A five-fold cross-validation study is performed to verify that the proposed method is robust to different training and test images. Ablation experiments further demonstrate the effectiveness of the proposed SASA neuron and CRED algorithm. Promotions of the average IoU individually utilizing the SASA and CRED module add up to the final promotion of the full model, indicating that the SASA and CRED modules contribute to the different stages of model and data in the training process.

참고문헌 (Reference)

1 Chul Min Yeum, "Visual data classification in post-event building reconnaissance" Elsevier BV 155 : 16-24, 2018

2 Xiangxiong Kong, "Vision-Based Fatigue Crack Detection of Steel Structures Using Video Feature Tracking" Wiley 33 (33): 783-799, 2018

3 Chul Min Yeum, "Vision-Based Automated Crack Detection for Bridge Inspection" Wiley 30 (30): 759-770, 2015

4 Simonyan, K., "Very deep convolutional networks for large-scale image recognition"

5 Mukkamala, M. C., "Variants of RMSprop and Adagrad with logarithmic regret bounds" 2545-2553, 2017

6 Abhinav Agrawal, "Using CNN for facial expression recognition: a study of the effects of kernel size and number of filters on accuracy" Springer Science and Business Media LLC 36 (36): 405-412, 2019

7 Zilong Wang, "Unsupervised deep learning approach using a deep auto-encoder with a one-class support vector machine to detect damage" SAGE Publications 20 (20): 406-425, 2020

8 Mohammad R. Jahanshahi, "Unsupervised Approach for Autonomous Pavement-Defect Detection and Quantification Using an Inexpensive Depth Sensor" American Society of Civil Engineers (ASCE) 27 (27): 743-754, 2013

9 Olaf Ronneberger, "U-Net: Convolutional Networks for Biomedical Image Segmentation" Springer International Publishing 234-241, 2015

10 Yuequan Bao, "The State of the Art of Data Science and Engineering in Structural Health Monitoring" Elsevier BV 5 (5): 234-242, 2019