Ring artifacts removal from synchrotron CT image slices

笛卡尔坐标系 极坐标系 傅里叶变换 小波变换 计算机视觉 人工智能 计算机科学 对角线的 光学 小波 物理 数学 几何学 数学分析
作者
Zhouping Wei,Sheldon Wiebe,Dean Chapman
出处
期刊:Journal of Instrumentation [IOP Publishing]
卷期号:8 (06): C06006-C06006 被引量:35
标识
DOI:10.1088/1748-0221/8/06/c06006
摘要

Ring artifacts can occur in reconstructed images from x-ray Computerized Tomography (CT) as full or partial concentric rings superimposed on the scanned structures. Due to the data corruption by those ring artifacts in CT images, qualitative and quantitative analysis of these images are compromised. In this paper, we propose to correct the ring artifacts on the reconstructed synchrotron radiation (SR) CT image slices. The proposed correction procedure includes the following steps: (1). transform the reconstructed CT images into polar coordinates; (2) apply discrete two-dimensional (2D) wavelet transform to the polar image to decompose it into four image components: low pass band image component, as well as the components from horizontal, vertical and diagonal details bands; (3). apply 2D Fourier transform to the vertical details band image component only, since the ring artifacts become vertical lines in the polar coordinates; (4). apply Gaussian filtering in Fourier domain along the abscissa direction to suppress the vertical lines, since the information of the vertical lines in Fourier domain is completely condensed to that direction; (5). perform inverse Fourier transform to get the corrected vertical details band image component; (6). perform inverse wavelet transform to get the corrected polar image; (7). transform the corrected polar image back to Cartesian coordinates to get the CT image slice with reduced ring artifacts. This approach has been successfully used on CT data acquired from the Biomedical Imaging and Therapy (BMIT) beamline in Canadian Light Source (CLS), and the results show that the ring artifacts in original SR CT images have been effectively suppressed with all the structure information in the image preserved.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
hey应助huzz采纳,获得20
刚刚
kukugogogo关注了科研通微信公众号
刚刚
刚刚
ywl发布了新的文献求助30
1秒前
lalalahei完成签到,获得积分10
1秒前
1秒前
科研通AI6.1应助薖上采纳,获得10
1秒前
1秒前
大仙完成签到,获得积分10
1秒前
超男发布了新的文献求助10
1秒前
潇潇木子完成签到,获得积分10
2秒前
科研通AI6.3应助zik采纳,获得30
2秒前
KK发布了新的文献求助10
2秒前
2秒前
2秒前
NexusExplorer应助柔弱的不尤采纳,获得10
2秒前
3秒前
zz完成签到,获得积分10
3秒前
liuying发布了新的文献求助20
3秒前
3秒前
3秒前
传奇3应助qq.com采纳,获得10
4秒前
天天向上发布了新的文献求助10
4秒前
液体剑0932发布了新的文献求助30
4秒前
56555完成签到,获得积分10
4秒前
qq发布了新的文献求助10
4秒前
万能图书馆应助hhh123采纳,获得10
5秒前
5秒前
文静山河应助芥子采纳,获得80
5秒前
lalalahei发布了新的文献求助10
5秒前
6秒前
科研通AI6.1应助小疯采纳,获得10
6秒前
sekiro发布了新的文献求助10
7秒前
徐海玲发布了新的文献求助10
7秒前
脑洞疼应助Ljr123采纳,获得10
7秒前
7秒前
7秒前
祖乐松发布了新的文献求助10
9秒前
9秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Kinesiophobia : a new view of chronic pain behavior 3000
Les Mantodea de guyane 2500
Molecular Biology of Cancer: Mechanisms, Targets, and Therapeutics 2000
What is the Future of Psychotherapy in a Digital Age? 700
The Psychological Quest for Meaning 600
Zeolites: From Fundamentals to Emerging Applications 600
热门求助领域 (近24小时)
化学 材料科学 生物 医学 工程类 计算机科学 有机化学 物理 生物化学 纳米技术 复合材料 内科学 化学工程 人工智能 催化作用 遗传学 数学 基因 量子力学 物理化学
热门帖子
关注 科研通微信公众号,转发送积分 5954812
求助须知:如何正确求助?哪些是违规求助? 7163630
关于积分的说明 15935734
捐赠科研通 5089659
什么是DOI,文献DOI怎么找? 2735382
邀请新用户注册赠送积分活动 1696186
关于科研通互助平台的介绍 1617224