A neural network ensemble method for effective crack segmentation using fully convolutional networks and multi-scale str
- PDF / 3,338,222 Bytes
- 18 Pages / 595.276 x 790.866 pts Page_size
- 22 Downloads / 190 Views
ORIGINAL PAPER
A neural network ensemble method for effective crack segmentation using fully convolutional networks and multi-scale structured forests Sen Wang1 · Xing Wu1,2
· Yinghui Zhang1 · Xiaoqin Liu1 · Lun Zhao3
Received: 16 September 2019 / Revised: 9 June 2020 / Accepted: 17 August 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Crack image segmentation has recently become a major research topic in nondestructive inspection. However, the image segmentation methods are not robust to variations such as illumination, weather, noise and the segmentation accuracy which cannot meet the requirements of practical applications. Therefore, a neural network ensemble method is proposed for effective crack segmentation in this paper, which consists of fully convolution networks (FCN) and multi-scale structured forests for edge detection (SFD). In order to improve the accuracy of crack segmentation and reduce the error mark under complex background, a new network model based on FCN model is proposed to address the problems that lose local information and the capacity of partial refinement, which are frequently encountered in FCN model in the crack segmentation. In addition, SFD is combined with the half-reconstruction method of anti-symmetrical bi-orthogonal wavelet to overcome the limitation of crack edge detection. Finally, the result of the two maps is merged after resizing to the original image dimensions. Qualitative and quantitative evaluations of the proposed methods are performed, showing that they can obtain better results than certain existing methods for crack segmentation. Keywords Crack image segmentation · Fully convolutional network · Multi-scale structured forests · Anti-symmetrical bi-orthogonal wavelet
1 Introduction Stress concentration and alternating load are inevitable accompanied with the manufacturing, transportation and application process, which may lead to harmful fatigue fracture during a long-term application. Due to different materials feature and complex object structure, various forms of external force and the larger dynamic range of the scene, the surface testing of the structure is a high difficulty work, which needs practical theory and has important social sig-
B
Xing Wu [email protected] Sen Wang [email protected]
1
Faculty of Mechanical and Electrical Engineering, Kunming University of Science and Technology, Yunnan 650500, China
2
Yunnan Vocational College of Mechanical and Electrical Technology, Yunnan 650203, China
3
School of Mechanical Engineering, Guizhou University, Guizhou 550025, China
nificance and economic value. Although the traditional crack testing technologies involved in different fields of social production and have unique advantage in various objects of detection, most methods require operation in accordance with the specific testing and installation conditions. The detection of a surface crack is usually performed by conventional human inspection, which is time-consuming, labor-intensive and subjective. Nondestructive insp
Data Loading...
