一种基于自动显微镜的X光图片快速图像拼接算法

摘要

为提高工业无损探伤中对X光图片的检测效率,提出用自动数字显微镜替代肉眼对X光图片进行自动检测,并通过研究自动显微镜图像采样方式和X光图像的特点,提出一种改进的基于空间像素灰度相关的快速图像拼接算法:以前一幅采样图像矩阵最后一列的中点为基点建立匹配矩阵,设立阈值,并通过相关性匹配找到下一采样图像中的匹配点,然后进行多重循环校验来验证匹配点的准确性.仿真实验证明,该算法不仅精确性高,而且运算量小,具有很高的运算效率和采样效率.%In order to improve the efficiency of X films detection in industry, a method using the automatic microscopy instead of the naked eyes to detect the X films is offered, and by researching the character of the microscopy sampling and the X film image, a fast approach of image stitching using gray relativity of pixels is proposed: taking out mid-pixel of the last row of frontal sampling image to construct the matching matrix, making the threshold, finding the matching pixel in the next image by getting the correlation, and validating the veracity by circle checkout. According to the experiment, this approach is of perfect accuracy and efficiency.