燃油喷雾内部构造可视化的激光CT技术(1)——喷雾图象重建方法的研究

Visualization of internal structure of fuel spray using laser computed tomography technique(1)——A study of image reconstruction algorithms for fuel spray

摘要: 为应用激光CT(Computed Tomography)技术实现燃油喷雾内部构造可视化,本文针对燃油喷雾的特点,对卷积反投影法(Convolution back projection简称CBP)、代数重建技术(Algebraic reconstruction technique简称ART)和同时迭代重建技术(Simultaneous iterative reconstruction technique简称SIRT)三种算法应用于喷雾图象的重建进行了研究.在P540图象处理系统上对标准喷雾模型进行重建,结果表明在投影数据存在噪声时,SIRT法重建图象质量最佳,但所需时间最长,CBP法重建图象质量最差,但运算时间最少,而ART法介于两者之间.

Abstract: According to the characteristics of fuel spray,CBP(Convolution back projection),ART (Algebraic reconstruction technique) and SIRT (Simultaneous iterative reconstruction technique) were studied on the image reconstruction of fuel spray.The reconstruction results of the phantom using P540 digital image processing system showed that when there is noise in the projection data,the best image was reconstructed by SIRT whose calculation time was longest,the worst image was reconstructed by CBP whose calculation time was shortest,and ART is between them.