基于数学形态学的图像预处理算法的研究

Study on Mathematical Morphology-based Image Preprocessing Algorithm

该文提出了一种基于数学形态学的图像预处理方法。首先对原始图像进行了改进的形态滤波处理,依次用一个略大于目标尺寸的结构元素和一个小于目标尺寸的结构元素对原始图像进行开运算;求两次滤波所得图像之差作为图像增强后的结果。实验结果表明,此方法最大限度地消除了各种噪声和畸变对图像中目标区域的影响,更加完整的保留了目标边缘的细节信息。

A novel image preprocessing method based on mathematical morphology is presented. First of all,the tyre cord fabric defects image is performed with the improved morphological filters. A frame element which is bigger than the target and a smaller one are used by turns in the opening operation;the subtract output is the image enhancement result. Then the improved morphological erode edge detection is used to reduce the edge information that erode detection dismissed. Final experiment results shows that the effect of different kinds of noise and aberrance on the target area is suppressed furthest,and more detail informations of the target are reserved.