磁瓦表面图像的下包络线灰度对比度缺陷检测算法
Defect Extraction on Magnetic Tile Surfaces based on Lower Envelope Gray-scale Contrast

为解决磁瓦表面缺陷对比度低、整体亮度不均匀以及磁瓦图像中存在大量的冲击噪声干扰等难题，提出了一种基于下包络线灰度对比度的缺陷检测算法。首先定义扫描线灰度对比度，用下包络的方式来优化每一行扫描灰度曲线，然后计算下包络线上的最大灰度对比度并判断该点是否为缺陷区域中的点从而得到缺陷区域的“骨架”，最后通过8邻接方式对“骨架”进行扩展得到完整的缺陷区域。实验证明，该方法能够有效解决冲击噪声对图像分割的影响，克服光照不均、表面存在磨削条纹等干扰，对不同类型缺陷有较好的分割效果。
In order to overcome low contrast and uneven brightness of magnetic tile surface image with much impulsive noise, a new segmentation method based on lower envelope gray-scale contrast was proposed in this paper. Firstly, the scanning-line gray-scale contrast is defined and every row scanning-line is optimized by envelope. Then the max gray-scale contrast is computed, the point in the defect area is confirmed and the "skeleton" of defect is got. Finally, the whole defect area by extending the "skeleton" with 8-adjacency connectivity is obtained. The experimental results show that the proposed method could obtain a good result of defect extraction in much impulse noise environment and fits on various defects, which lays a good foundation for subsequent feature extraction and recognition classification of defects