Japanese Geotechnical Society Special Publication
Online ISSN : 2188-8027
ISSN-L : 2188-8027
14. Innovative technologies and informatics
Enhancing images for particle image velocimetry in centrifuge models
Kai Qi TanQiao Yue TungFook Hou LeeSiang Huat Goh
著者情報
ジャーナル フリー

2016 年 2 巻 74 号 p. 2553-2558

詳細
抄録
Particle image velocimetry (PIV) is a widely used image processing technique for mapping deformation in centrifuge models. One of the problems associated with PIV is the occurrence of incorrect or ‘ wild vectors’. A ‘ wild vector’ is defined as a vector that does not follow the general directional trend of the other vectors surrounding it, which if present in large numbers, can affect the accuracy of the deformation plot. This problem is particularly important in clay model, the surfaces of which often do not have inherent texture. The paper examines some possible enhancements to PIV images by fine tuning three parameters, namely type of texture material, method of texture placement and the use of high resolution colour imagery. The trials were conducted on centrifuge models of tunnels. The desired outcome is to minimize the percentage of wild vectors that is attained from the PIV analyses of the recorded images. The results show that with the use of colour model terrain, the likelihood of clay obscuration can be reduced significantly, Furthermore, by the use of an appropriate method of scatter will offer favorable image contrast for PIV based experiments involving the use of saturated clay. Finally the advantage of colour imagery is that it is able to capture the difference in color as the coloured camera is able to detect the hue, saturation and luminance of the texture used. These lead to a corresponding improvement in image contrast and hence, PIV analysis results.
著者関連情報
前の記事 次の記事
feedback
Top