小波分析在织物缝纫平整度客观评价中的应用

Application of wavelet analysis in evaluation on seam pucker of fabrics

提出一种基于小波分析的缝纫平整度客观评价方法,可以克服主观评价的局限性,使得缝纫外观质量的检测更加快速、准确。在对缝纫平整度标样图像进行处理的基础上,采用二维离散小波变换提取图像高频分量中细节系数的标准差作为特征向量,其在特定的分析尺度上随缝纫平整度的恶化呈现出明显的单调递增趋势。通过计算平整度评估因子(EK)确定了最佳小波和最佳小波分析尺度,从而建立了缝纫平整度等级的客观评判模型。采用棉织物的缝纫样本检验了模型,结果表明主客观评价的吻合度达到85%以上,由此验证了该方法的可行性。

A new objective method based on wavelet analysis is put forward to which can overcome the traditional limitations of subjective evaluation, making evaluate seam pucker of fabric, the detect of sewing appearance more rapid and accurate. Firstly, the standard images to evaluate seam pucker are processed and transformed by 2-D discrete wavelet analysis. Then, the standard deviations of detail coefficient on the high frequent, as an eigenvalue, are picked up. By computing evaluation index （EK）, the optimal dimension and optimal wavelet are confirmed. On this basis, the objective evaluation model is established. Finally, this model is validated by some sewed cotton samples, and the results show that the match degree of subjective and objective evaluation is 85% or more, which verify the feasibility of the method.