基于多尺度效应的坯布缺陷边缘检测算法

Edge Detection Algorithm Based on Multi-scale Effect of Fabric Defects

生产线上运动产品缺陷图像的边缘具有模糊和不确定性,采用多尺度方法来描述和分析可以获得更多信息。为了满足当前坯布检验过程中对于产品表面缺陷进行快速准确检测的要求,文中提出了一种基于多尺度效应的图像边缘检测与分割方法,结合小波变换的图像频带分解与重构、粗糙集的图像区域分割增强、多结构元素的形态学边缘提取方法为一体,应用于坯布表面质量检测,取得了较为理想的分割效果。

The edges of motion product＇s defect image are often with ambiguity and uncertainty, a muhi-scale method to describe and analyze is of more information. In order to meet the fast and accurate detection requirements for the product surface defects in current fabric inspection process. An image edge detection and segmentation method was proposed combined with imageband decomposition and reconstruction based on Wavelet transform, image region segmentation and enhancement based on Rough Set, and multi-structure element morphological edge extraction. It has been applies to the fabric surface quality inspection and obtained the ideal segmentation effect.