Fabric Patterns Recognition Based on Weft Phase Difference

For the purpose of identifying yarn density and fabric weave patterns, a novel method for fabric patterns recognition is proposed. The methods of gray-level integral projection and discrete Fourier transform are used to estimate yarn density in this paper. To avoid errors caused by noise, yarn deformation and illumination,yarn statistics are used. The concept of the weft phase difference is proposed which is the key point to derive the fabric patterns. Finally, a pattern with minimum error is identified. Two experiments are used to evaluate the proposed method. It is shown that our method is effective and robust. For some other methods, it is ineffective for large-scale repeat unit satin fabrics. However, our approach is still effective.