Improved capacitance imaging biconjugate gradient image reconstruction algorithm
-
摘要:
针对电容层析成像技术(ECT)逆问题中软场效应的影响,以及重建图像时使用的传统迭代类算法迭代次数多、成像速度慢等问题,将双共轭梯度(BICG)应用到电容层析成像技术中,为了得到更好的重建效果,提出了双共轭梯度与正则化思想相结合来求解逆问题的最佳解。通过COMSOL5.3软件进行建模,使用MATLAB 2014a进行图像重建与图像评估,分别使用Tikhonov、Landweber、共轭梯度(CG)、BICG、所提改进算法进行图像重建。实验表明:所提改进算法的成像效果不仅优于其他迭代类算法,而且大大缩短了图像重建需要的时间;尤其对一些复杂流型成像效果更佳,图像错误率低至约0.2,相关系数高达约0.88,成像时间缩短至2.77 s,迭代次数减少至20次。
-
关键词:
- 电容层析成像技术(ECT) /
- 双共轭梯度(BICG)算法 /
- 算法改进 /
- 图像重建 /
- COMSOL
Abstract:Aimed at the influence of the soft field effect on the inverse electrical capacitance tomography (ECT) problem and to solve the problems that the traditional iterative algorithm used to reconstruct images has numerous iteration times, and that the imaging speed is slow, for the first time, biconjugate gradient (BICG) is applied to the capacitance tomography technology. In order to obtain a better reconstruction effect, a BICG and regularization idea are combined to solve the optimal solution of the inverse problem. The experiment was modeled by COMSOL5.3 software and MATLAB 2014a was used for image reconstruction and image evaluation. The images were reconstructed using Tikhonov, Landweber, conjugate gradient (CG), BICG and improved biconjugate gradient algorithm. Experiments show that the effect of the improved biconjugate gradient algorithm is not only better than other iterative algorithms, but also greatly reduces the time required for image reconstruction. The experimental results show that the biconjugate gradient algorithm has better imaging effect in some complicated flow patterns. The error is as low as 0.2, the correlation coefficient is as high as 0.88, the imaging time is reduced to 2.77 s, and the number of iterations is reduced to 20 times.
-
表 1 IME计算值
Table 1. Calculated values of IME
仿真模型 Tikhonov算法 Landweber算法 CG算法 BICG算法 改进算法 核心流 0.3880 0.2537 0.2420 0.2857 0.2524 双泡流 0.4201 0.3813 0.4227 0.3512 0.3503 三泡流 0.4802 0.4254 0.5012 0.4003 0.4010 层流 0.5046 0.4213 0.9344 0.3612 0.3410 环流 0.6548 0.5826 0.5802 0.2013 0.2106 表 2 CORR计算值
Table 2. Calculated values of CORR
仿真模型 Tikhonov算法 Landweber算法 CG算法 BICG算法 改进算法 核心流 0.7012 0.8063 0.7901 0.7886 0.8095 双泡流 0.6325 0.6418 0.6558 0.7802 0.7795 三泡流 0.5361 0.5904 0.5418 0.6217 0.6223 层流 0.5241 0.6908 0.0130 0.7146 0.7134 环流 0.4529 0.4608 0.4621 0.8592 0.8786 表 3 成像算法数据对比
Table 3. Comparison of imaging algorithm data
算法 迭代次数 成像时间/s Tikhonov算法 3.06 Landweber算法 120 4.47 CG算法 200 10.32 BICG算法 50 3.12 改进算法 20 2.77 -
[1] 张立峰.压缩感知在电容层析成像中的应用[J].北京航空航天大学学报, 2017, 43(11):2316-2321. https://bhxb.buaa.edu.cn/CN/abstract/abstract14344.shtmlZHANG L F.Application of compressed sensing in electrical capacitance tomography[J].Journal of Beijing University of Aeronautics and Astronautics, 2017, 43(11):2316-2321(in Chinese). https://bhxb.buaa.edu.cn/CN/abstract/abstract14344.shtml [2] BROWN B E.International society for reef studies[J].Coral Reefs, 1984, 3(1):58. doi: 10.1007/BF00306141 [3] KAGAWA Y, MURAI T.Numerical simulation of acoustic levitation[C]//IEEE 1985 Ultrasonics Symposium.Piscataway, NJ: IEEE Press, 1985: 198574. [4] VAUHKONEN M, KARJALAINEN P A, KAIPIO J P.A Kalman filter approach applied to the tracking of fast movements of organ boundaries[C]//Proceedings of the 20th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.Piscataway, NJ: IEEE Press, 1998: 6345874. [5] YANG W Q, SPINK D M, YORK T A, et al.An image reconstruction algorithm based on Landweber's iteration method for electrical capacitance tomography[J].Measurement Science and Technology, 1999, 10(11):1065-1069. doi: 10.1088/0957-0233/10/11/315 [6] KIM D K, KIM S B, KIM H K.Image processed tracking system of multiple moving objects based on Kalman filter[J].KSME International Journal, 2002, 16(4):427-435. doi: 10.1007/BF03185072 [7] YANG W Q, STOTT A L, GAMIO J C.Analysis of the effect of stray capacitance on AC-based capacitance transducer for tomography[J].IEEE Transactions on Instrumentation and Measurement, 2003, 52(5):1674-1681. doi: 10.1109/TIM.2003.817925 [8] 王化祥, 朱学明, 张立峰.用于电容层析成像技术的共轭梯度算法[J].天津大学学报, 2005, 38(1):1-4. http://d.old.wanfangdata.com.cn/Periodical/tianjdxxb200501001WANG H X, ZHU X M, ZHANG L F.Conjugate gradient algorithm for electrical capacitance tomography[J].Journal of Tianjin University, 2005, 38(1):1-4(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/tianjdxxb200501001 [9] 王化祥, 胡理, 赵波.基于自适应网格剖分的EIT图像重建算法[C]//中国生物医学工程进展——2007中国生物医学工程联合学术年会, 2007: 5.WANG H X, HU L, ZHAO B.EIT image reconstruction algorithm based on adaptive meshing[C]//Progress in Biomedical Engineering in China—2007 China Biomedical Engineering Joint Academic Annual Meeting, 2007: 5(in Chinese). [10] 田沛, 秦京建, 秦旭刚.基于两相流电容层析成像系统的正则化Landweber算法[J].仪器仪表用户, 2018, 25(9):90-92. doi: 10.3969/j.issn.1671-1041.2018.09.027TIAN P, QIN J J, QIN X G.Regularized Landweber algorithm based on two-phase electrical capacitance tomography system[J].Instrumentation Users, 2018, 25(9):90-92(in Chinese). doi: 10.3969/j.issn.1671-1041.2018.09.027 [11] 袁俊峰.基于并行稳定双共轭梯度算法的不可压缩管流数值模拟[D].杭州: 杭州电子科技大学, 2013.YUAN J F.Numerical simulation of incompressible tube flow based on parallel stable double conjugate gradient algorithm[D].Hangzhou: Hangzhou University of Electronic Science and Technology, 2013(in Chinese). [12] 彭政, 陈东方, 王晓峰.正则化超分辨率重建过程的自适应阈值去噪[J].计算机应用, 2017, 37(7):2084-2088. doi: 10.3969/j.issn.1001-3695.2017.07.036PENG Z, CHEN D F, WANG X F.Adaptive threshold denoising of regularized super-resolution reconstruction process[J].Journal of Computer Applications, 2017, 37(7):2084-2088(in Chinese). doi: 10.3969/j.issn.1001-3695.2017.07.036 [13] 陈德运, 陈宇, 王莉莉, 等.基于改进Gauss-Newton的电容层析成像图像重建算法[J].电子学报, 2009, 37(4):739-743. doi: 10.3321/j.issn:0372-2112.2009.04.012CHEN D Y, CHEN Y, WANG L L, et al.Image reconstruction algorithm based on improved Gauss-Newton for electrical capacitance tomography[J].Chinese Journal of Electronics, 2009, 37(4):739-743(in Chinese). doi: 10.3321/j.issn:0372-2112.2009.04.012 [14] CUI Z, WANG H, YIN W.Electrical capacitance tomography with differential sensor[J].IEEE Sensors Journal, 2015, 15(9):5087-5094. doi: 10.1109/JSEN.2015.2446982 [15] 马平, 司志宁.基于小波变换的CT/ECT图像融合方法[J].计量学报, 2018, 39(4):536-540. doi: 10.3969/j.issn.1000-1158.2018.04.18MA P, SI Z N.CT/ECT image fusion method based on wavelet transform[J].Acta Metrologica Sinica, 2018, 39(4):536-540(in Chinese). doi: 10.3969/j.issn.1000-1158.2018.04.18