Premium
Quantitative image processing analysis of composite materials
Author(s) -
Yang Heechun,
Colton Jonathan S.
Publication year - 1994
Publication title -
polymer composites
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.577
H-Index - 82
eISSN - 1548-0569
pISSN - 0272-8397
DOI - 10.1002/pc.750150108
Subject(s) - thresholding , materials science , orientation (vector space) , image processing , enhanced data rates for gsm evolution , artificial intelligence , fiber , thermoplastic , thermoplastic composites , image (mathematics) , computer vision , composite material , computer science , geometry , mathematics
Quantitative digital image processing has been used to analyze unidirectional thermoplastic composites. The fiber volume fraction is determined by the template matching method and compared with the conventional thresholding method. A combination of skeleton and edge cutting is used to obtain the centers of the fibred in the image. The Voronoi distances are calculated from the result. The Hough transform is applied to the fiber orientation in conjunction with thresholding by the line length in the image. To describe the fiber orientation, characteristic function are defined in the width and thickness directions, which allow a much more accurate and effective description of the fiber distribution and orientation than other methods. Advantages of the template method are that similar results to the threshold method can be obtained as thresholding with a lower magnifications, resulting in an improved statistical sample with a reduced number of photomicro‐graphs.