基于傅里叶变换的随机纹理瓷砖的高精度表面缺陷检测

High-precisionSSurface Defect Detection of Random Texture Ceramic Tiles using Fourier Transform

摘要:

针对经典缺陷检测算法不能很好地提取随机纹理瓷砖图像的缺陷的问题,提出一种基于傅里叶变换的随机纹理瓷砖表面缺陷高精度检测方法。在此基础上,完成了瓷砖表面缺陷检测硬件系统设计。对采集的瓷砖图像,首先利用傅里叶变换得到频率谱图像,然后研究截止频率参数对滤波的影响,设计最优化滤波器进行滤波,再通过傅里叶逆变换获得重构图像,达到抑制背景纹理信息,加强缺陷区域信息的目的,最后通过阈值化和形态学操作获得缺陷区域。实验结果表明: 本方法对不同的随机纹理瓷砖样本进行缺陷检测的准确率高,在瓷砖缺陷检测中具有较高的实用价值。

Abstract:

Aiming at the problem that the defects of random texture tile images can not be well extracted via the classical defect detection algorithm,a high-precisionSsurface defect detection method of random texture ceramic tiles using Fourier Transform is proposed.On this basis, the hardware system of ceramic tile surface defect detection is designed.Firt frequency spectrum images are obtained by Fourier transform for the collected tile images, then the influence of cut-off frequency parameter on filtering is studied and the optimal filter is designed for filtering, and then reconstructed images are obtained through the inverse discrete Fourier transform to suppressing background texture information and strengthening the defect area information,finally,the defect areas are acquired by threshold and morphological operation.Experimental results show that this method has high accuracy on defect detection for different random texture ceramic tile samples, and it has high practical value in ceramic tile defect detection.