X射线工业CT成像过程复杂伪影抑制方法综述  

Review of Complex Artifact Reduction Methods for Industrial Computerized Tomography Imaging

作　　者：杨富强[1,2] 杨瑶 李志翔 黄魁东[1,2] YANG Fu-Qiang;YANG Yao;LI Zhi-Xiang;HUANG Kui-Dong(Key Laboratory of High Performance Manufacturing for Aero Engine,Ministry of Industry and Information Technology,Northwestern Polytechnical University,Xi'an 710072;Engineering Research Center of Advanced Manufacturing Technology for Aero Engine,Ministry of Education,Northwestern Polytechnical University,Xi'an 710072)

机构地区：[1]西北工业大学航空发动机高性能制造工业和信息化部重点实验室,西安710072 [2]西北工业大学航空发动机先进制造技术教育部工程研究中心,西安710072

出　　处：《自动化学报》2023年第4期687-704,共18页Acta Automatica Sinica

基　　金：国家自然科学基金青年基金(52005415);国家科技重大专项(J2019-Ⅶ-0013-0153);航空发动机及燃气轮机基础科学中心项目(P2022-B-IV-013-001);中国航空发动机集团产学研合作项目(HFZL2022CX Y024);中央高校基本科研业务费专项资金(HXGJXM202216)资助。

摘　　要：X射线工业计算机断层(Computerized tomography,CT)技术是一种先进的非接触式无损三维检测技术,能在无损伤情况下以灰度图像的形式对物体内部结构进行全面、详细地分析,在航空航天、工业生产、安检等领域发挥着重要的作用.针对工业CT伪影严重降低图像质量问题,对工业CT成像过程复杂伪影形成机理进行分析,对不同类型伪影抑制方法进行归纳总结.阐述了基于射线衰减、探测器及高密度差异、采样数据及重建等不同过程伪影成因及伪影消除相关算法的最新技术进展,并对近年来人工智能深度学习背景下新兴的基于深度学习及神经网络的工业CT无损检测研究与发展方向进行了总结和展望.Computerized tomography(CT)technology is an advanced non-destructive three-dimensional detection technology that can completely investigate the interior structure of an item in the form of grayscale images without contacting.It plays an important role in aerospace,industrial production,security inspection and other fields.Because industrial CT artifacts have had a significant impact on image resolution,this study examines the mechanism of coupling artifacts in the industrial CT imaging process,as well as the most recent technical advancements in artifact suppression methods.For the artifacts created during the ray attenuation process,the item containing scattering artifacts and cupping artifacts is described first.Second,the related technologies for artifact generated by the detector and the object suppression,such as ring artifacts and metal artifacts are discussed.And then,the latest developments in the suppression of streaking artifacts,banding artifacts and blocking artifacts are presented for the accessories generated by sampling data and reconstruction methods.By reviewing the research of industrial CT artifact suppression in recent years,the challenges and the development in the field of high-resolution imaging based on artificial intelligence in deep learning are summarized and prospected.

关 键 词：工业CT 伪影抑制 深度学习 智能检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]