为在卷烟爆珠放行检验中准确检测出气泡缺陷爆珠,避免将污点、白斑以及胶皮粘附等非必检缺陷误识别为气泡,基于计算机视觉技术建立了一种高精度气泡缺陷检测方法。首先,在HSV(Hue,Saturation,Value)颜色空间中对图像进行分割,实现爆珠区域的精确提取;其次,对LoG(Laplacian of Gaussian)边缘检测算子进行改进,增强气泡轮廓,抑制其他缺陷轮廓;最后,根据气泡的“双液环”特性对提取出的轮廓进行编码,准确识别出气泡。采用3种不同规格、不同颜色的爆珠样品对该方法进行测试,结果表明:本文方法对气泡缺陷的平均识别率为98.9%,平均误检率为0.21‰,与人工检测对比偏差为±0.03%,对不同类别爆珠的气泡缺陷均具有较高识别率。该方法可为提高卷烟爆珠外观质量检测水平提供支持。
The polydimethylsilicone(PDMS) rubber, a material generally used to make the elastomeric stamp in soft lithography, was modified by the organic montmorillonite(MMT). The cure behavior and the modular change of the nanocomposite were monitored with the resin curemeter. The results showed that after the intercalation of montmorillonite, the nanocomposite cure faster and the modulus of compound increased obviously. Especially, the elasticity and fitness of the PDMS/MMT nanocomposite were enhanced, also the nanocomposite had better mechanical properties and stronger solvent resistance. The results suggest that this nanocomposites is more suitable for preparing elastomeric stamps in soft lithography.
