基于预浸纱自动铺放缺陷的分割算法

为保证预浸纱自动铺放成型制件的性能,基于图像处理技术,根据预浸纱表面图像沿纤维方向灰度均匀的特性,提出了一种结合灰度补偿和差影分割的缺陷分割算法。利用图像灰度补偿矩阵对图像进行灰度补偿,依据图像的灰度误差服从正态分布的特点剔除图像矩阵中的极大误差点,并建立标准背景图像,以允许差值系数K作为判据进行差影分割。结果表明:该算法对气泡、异物两种典型铺放缺陷分割效果好、速度快,为实现自动铺放缺陷在线监测提供了良好的理论基础。

Algorithm of Defect Segmentation for AFP Based on Prepregs

In order to ensure the performance of the automated fiber placement forming parts,according to the homogeneity of the image of the prepreg surface along the fiber direction,a defect segmentation algorithm which was the combination of gray compensation and substraction algorithm based on image processing technology was proposed.The gray compensation matrix of image was used to compensate the gray image,and the maximum error point of the image matrix was eliminated according to the characteristics that the gray error obeys the normal distribution.The standard image was established,using the allowed deviation coefficient K as a criterion for substraction segmentation.Experiments show that the algorithm has good effect,fast speed in segmenting two kinds of typical laying defect of bubbles or foreign objects,and provides a good theoretical basis to realize automatic laying defect online monitoring.