基于人工智能的光学表面疵病图像自动化拼接研究  

Research on Automated Splicing of Optical Surface Defect Images Based on Artificial Intelligence

作　　者：何佳 HE Jia(School of Electronic and Information Engineering,Xi’an Siyuan University,Xi’an 710038,China)

机构地区：[1]西安思源学院电子信息工程学院,西安710038

出　　处：《自动化与仪表》2024年第9期61-65,共5页Automation & Instrumentation

摘　　要：常规方法图像自动化拼接质量较差,为此提出基于人工智能的光学表面疵病图像自动化拼接方法。光学元件表面疵病检测系统对采集图像进行尺度缩放与向上采样,求取图像序列矩阵,结合重叠区域的交线权值提取图像基准特征点,并采用人工智能算法配准图像,对基准特征点进行匹配,以此为依据,通过融合多频段图像与分析图像梯度的敏感度变化规律对待拼接图像的拼接区域进行标记、实现图像自动化拼接。以光学表面疵病图像为实验对象,实验结果表明,所提方法在不同测试图像中获得的自动化拼接图像自然度均在1.0以上,自动化拼接图像自然度较高,图像的自动化拼接质量较好。The quality of automated image stitching using conventional methods is poor.Therefore,an artificial intelligence based automated stitching method for optical surface defect images is proposed.The optical component surface defect detection system scales and upsamples the collected images,calculates the image sequence matrix,extracts image reference feature points based on the intersection weight of overlapping areas,and uses artificial intelligence algorithms to register the images.The reference feature points are matched based on this.By fusing multi frequency band images and analyzing the sensitivity change law of image gradients,the automated stitching area of the image is marked to achieve automatic image stitching.Taking optical surface defect images as the experimental object,the experimental results show that the proposed method obtains automated stitching images with naturalness above 1.0 in different test images.The naturalness of automated stitching images is high,and the quality of automated stitching images is good.

关 键 词：人工智能 光学元件 表面疵病图像 图像自动化拼接 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391[自动化与计算机技术—控制科学与工程]