基于形态学和小波变换的瓷砖缺陷边缘检测

Tile Defects Detection Based on Morphology and Wavelet Transformation

针对传统算法难以检测复杂的不规则花纹瓷砖缺陷的问题,提出一种基于形态学算子和小波变换融合的边缘检测算法。上述算法利用了小波变换能够把图像高低频分离、重构的特性,在经过二层小波变换后,用设计的自适应形态学算子对低频图像中的缺陷、花纹及背景分离;将提出的高通滤波器用于高频轮廓细节信息的提取,重构得到检测的缺陷轮廓。经多组实验对比仿真证明,与通用方法相比,提出的算法能有效地提取复杂背景缺陷轮廓,对瓷砖及其它产品的边缘检测有积极的作用。

Due to the traditional algorithms are difficult to detect complex irregular tile defects,an edge detection algorithm based on morphology and wavelet transform fusion is proposed. The algorithm utilizes the wavelet transformation to separate image from high to low frequency,after the secondary wavelet transformation,the design of the adaptive morphological operator is used to separate the defects from the pattern background in the low frequency image,and the proposed high-pass filter is used for the extraction of high-frequency detail information. The simulation results show that,compared with the general method,the proposed algorithm can effectively extract the contours of complex background defects and has a positive effect on the edge detection of tiles and other products.