摘要
针对计算层析成像火焰三维重建,本文发展了一种新的迭代算法,乘型代数重建非局部平均滤波算法(MART-NLM),通过与现有迭代算法进行对比,发现MART-NLM具有更好的重建精度,可以有效抑制噪声和伪影,改善火焰锋面的重建质量;在存在5°角度误差及5%高斯噪声下仍能重建出火焰轮廓,具有较好鲁棒性.
Aiming at the 3D reconstruction of flames based on computed tomography,this work proposes a new algorithm:the multiplicative algebraic reconstruction technique with non-local means(MART-NLM)filtering algorithm.By comparing it with existing algorithms,it is found that MART-NLM has better reconstruction accuracy,can effectively suppress noise and artifacts,and improve the reconstruction quality of flame front.Even with a 5°angle error and 5%Gaussian noise,the flame contour can still be reconstructed with good robustness.
作者
李响
雷庆春
徐文江
范玮
Li Xiang;Lei Qingchun;Xu Wenjiang;Fan Wei(School of Power and Energy,Northwestern Polytechnical University,Xi’an 710129,China;School of Aerospace Engineering,Xiamen University,Xiamen 361005,China)
出处
《燃烧科学与技术》
CAS
CSCD
北大核心
2023年第6期660-666,共7页
Journal of Combustion Science and Technology
基金
国家科技重大专项资助项目(HT-J2019-I-0003-0004).
关键词
计算层析
火焰三维重建
乘型代数重建算法
computed tomography
flame 3D reconstruction
multiplicative algebraic reconstruction