X-Ray CT Images Segmentation for Asphalt Concrete Based on Fuzzy Network Algorithm

Identification of optimum aggregate gradation is important for pavement quality control and quality assurance of pavement construction. Traditional gradation detection methods such as chemical reagent extraction and burning asphalt mixture are environmentally unfriendly methods. X-ray computed tomography (CT) is an non-destructive technique for capturing the microstructure images of asphalt mixture. However, in order to improve visualization of the CT images, an accuracy digital image processing method should be developed . In this study, fuzzy network, multilevel thresholding and morphological methods are utilized to reduce the noise, to enhance the contrast and segment images. On the basis of the processed images, the 2D aggregate gradation obtained from the digital image segmentation procedure is transferred into 3D gradation by stereological method. The results show that fuzzy network based on the local image gray level information can balance the noise reduction and contrast enhancement. The multilevel Otsu’s thresholding method can obtain the two threshold automatically and hence it can eliminate artificial threshold selection error. The morphological processing and watershed transformation can fill the holes in the image and break the aggregate connections which come from noise and heterogeneous density of the mixture. By comparing the calculated gradation and the designed gradation, the image processing procedure proposed in this paper can be utilized to obtain the gradation information of asphalt concrete (AC), stone mastic asphalt (SMA) and open-graded fiction course (OGFC) accurately and effectively.

• Supplemental Notes:
  • This paper was sponsored by TRB committee AFK50 Standing Committee on Structural Requirements of Asphalt Mixtures.
• Authors:
  • Xing, Chao
  • Tan, Yiqiu
  • Liu, Xueyan
  • Zhou, Changhong
  • Scarpas, Tom
• Conference:
  • Transportation Research Board 97th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2018-1-7 to 2018-1-11
• Date: 2018

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 5p

Subject/Index Terms

• TRT Terms: Asphalt concrete pavements; Fuzzy algorithms; Image processing; Microstructure; Nondestructive tests; Tomography; X ray analysis
• Uncontrolled Terms: X-ray computed tomography
• Subject Areas: Highways; Materials; Pavements;

Filing Info

• Accession Number: 01661030
• Record Type: Publication
• Report/Paper Numbers: 18-01938
• Files: TRIS, TRB, ATRI
• Created Date: Feb 23 2018 4:28PM