Kaiser窗在滤波反投影图像重建中的应用研究  

Application of Kaiser Window in Filtered Back Projection Image Reconstruction

作　　者：于尚民 YU Shangmin(School of Electronic and Information,Qingdao University,Qingdao 266071)

机构地区：[1]青岛大学电子信息学院,青岛266071

出　　处：《计算机与数字工程》2020年第2期423-427,共5页Computer & Digital Engineering

摘　　要：计算机断层成像技术的核心是图像重建算法。重建算法直接影响重建速度和重建精度。滤波反投影(FBP)算法是一种经典解析类CT图像重建算法,其时间复杂度低,易于通过硬件并行计算,并能获得较高的图像重建精度。滤波反投影算法的关键在于滤波函数的选择,滤波函数对重建图像的精度至关重要。窗函数法常用于滤波函数的设计。Kaiser窗作为一种参数可调的窗函数有着广泛的应用空间。首先对基于经典窗函数的R-L和S-L滤波函数进行分析,在此基础上将Kaiser窗滤波函数和R-L滤波函数、S-L滤波函数进行对比,并分析了Kaiser窗滤波函数在泊松噪声下的图像重建的精度。通过计算机仿真实验的方法,验证了Kaiser窗滤波函数在具有较高图像重建精度的同时,还具有较强的抗噪能力。The core of computed tomography is image reconstruction algorithm. Reconstruction algorithm directly affects reconstruction speed and reconstruction accuracy. Filtered back projection(FBP)algorithm is a classical analytic CT image reconstruction algorithm. Its time complexity is low,so it is easy to use hardware to realize parallel computing. At the same time,it can obtain high accuracy of reconstructed image. The key of filtered back projection algorithm is the selection of filter function. Filtering function is crucial to the accuracy of reconstructed image. The window function method is often used in the design of filter function. As a window function with adjustable parameters,Kaiser window has wide application space. Firstly,this paper analyzes the R-L and S-L filtering functions which are based on the traditional window function. On this basis,this paper compares the filter function of the Kaiser window with the R-L filter function and the S-L filter function. Finally,this paper analyzes the accuracy of image reconstruction of Kaiser window filter function under Poisson noise. Through the computer simulation experiment,it is verified that the Kaiser window filter function has higher image reconstruction accuracy and stronger noise immunity.

关 键 词：KAISER窗 反投影算法 滤波函数 图像重建 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]