TY - JOUR
T1 - Automatic Measurement Method of Yarn Dyed Woven Fabric Density via Wavelet Transform Fusion Technique
AU - Rui Zhang & Binjie Xin
JO - Journal of Fiber Bioengineering and Informatics
VL - 2
SP - 115
EP - 132
PY - 2016
DA - 2016/05
SN - 9
DO - http://doi.org/10.3993/jfbim00236
UR - https://global-sci.org/intro/article_detail/jfbi/10596.html
KW - Density Measurement
KW - Yarn Dyed Woven Fabric
KW - Image Fusion
KW - Image Processing
KW - Dual-side Scanning
AB - The yarn density is always considered as the fundamental structural parameter used for the quality
evaluation of woven fabrics. Conventional yarn density is measured by one-side image analysis. In this
paper, a novel density measurement method is developed for yarn dyed woven fabrics based on dual-side
fusion technique. Firstly, both face-side and back-side image of woven fabric are acquired via a lab-used
dual-side imaging system. Secondly, the affine transform is used for dual-side image alignment. Then, the
color images of woven fabrics are transferred from RGB to CIE-Lab color space, the intensity information
of image extracted from L component is used for the texture fusion and analysis. Subsequently, the dual-
side image is merged by the self-developed Wavelet transform fusion method. Finally, the fast Fourier
transform is utilized to measure yarn density of the fused image, the yarn alignment image could be
reconstructed using inverse fast Fourier transform. Our experimental results show that the accuracy of
density measurement by using proposed method is close to 99.44% compared with traditional method
and the robustness of this new proposed method is better than that of conventional analysis methods.