基于阈值分割和形态学相结合的金属废料 X 射线图像轮廓提取方法

针对金属废料 X 射线透射图像中废料之间距离较近时图像可能相互粘连,导致不能有效提取单个物料轮廓的问题,提出了一种基于阈值分割和形态学相结合的金属废料 X 射线图像轮廓提取方法。首先分析图像噪声信息分布,利用约束灰度值范围的 Otsu 阈值分割算法获得全局最优阈值对图像进行滤波;然后采用动态阈值分割方法得到二值化图像,在此基础上采取一种改进的形态学处理策略实现对各单物料区域的连通区标记;最后对各连通区进行轮廓提取以获得单个物料区域的轮廓。经实验表明,该方法有效地提高了在物料粘结情况下图像轮廓提取的准确率和成功率。

A Method for Extracting X-ray Image Contours of Metal Scraps Based on the Combination of Threshold Segmentation and Morphology

In the X-ray transmission image of metal scrap , when the distance between the scraps is too close , different scrap areas in the image may stick to each other , resulting in the inability to effectively extract the outline of a single scrap.Aiming at this problem , a method for extracting the contour of metal scrap X-ray images based on the combination of threshold segmentation and morphology is proposed. First the image noise information distribution is analyzed , the Otsu threshold segmentation algorithm that restricts the gray value range is used to obtain the global optimal threshold to filter the image , and then the dynamic threshold segmentation method is used to obtain the binarized image.On this basis , an improved morphological processing strategy is adopted to realize the marking of the connected area of each single waste material area.Finally , the contours of each connected area are extracted to obtain the contours of a single waste material area.Experimental results show that this method effectively improves the accuracy and success rate of image contour extraction when metal scraps are stuck to each other.