用快速哈达玛变换加速滤波反投影算法的滤波过程

乔志伟; 韩焱; 魏学业

doi:10.3724/SP.J.1146.2009.01635

用快速哈达玛变换加速滤波反投影算法的滤波过程

doi: 10.3724/SP.J.1146.2009.01635

乔志伟^{①② 韩焱,},
韩焱,
魏学业

基金项目:

国家自然科学基金(60602041，60532080)，山西省自然科学基金(2007012003)，电子测试技术国防科技重点实验室基金项目(9140C1204040908)和山西省高校科技项目(20091020)资助课题

计量
- 文章访问数: 3871
- HTML全文浏览量: 88
- PDF下载量: 1000
- 被引次数: 0
出版历程
- 收稿日期: 2009-12-23
- 修回日期: 2010-06-09
- 刊出日期: 2010-09-19

Accelerate the Filtering Process of Filtered Back Projection Algorithm Using Fast Hadamard Transform

Qiao Zhi-Wei^{①② 韩焱
,},
Han Yan,
Wei Xue-Ye

摘要

摘要: 为了加速滤波反投影算法的滤波过程，提出了用快速哈达玛变换(FHT)实现线性卷积的快速算法。分析了哈达玛变换的特点和快速算法的时间复杂度，设计了用FHT计算线性卷积的矩阵表达式，并推导出了哈达玛域滤波器的增益矩阵的求解公式，分析了该方法的加速原理及其适用条件。理论分析表明，该方法比FFT线性卷积法快了一倍。仿真实验表明，该方法在不影响图像重建精度的情况下，相对于FFT线性卷积法，将滤波过程的速度提高了近一倍。
- 图像重建 /
- 滤波反投影 /
- 加速 /
- 哈达玛变换 /
- 卷积
Abstract: To speedup the filtering process of the Filtered Back Projection (FBP) algorithm, a fast linear convolution algorithm using Fast Hadamard Transform (FHT) is proposed. The characteristic of Hadamard transform and its time complexity is analyzed. The matrix expression of using FHT to calculate linear convolution is designed. The equation of the Hadamard filter gain matrix is deduced. The acceleration principle and applicable condition of the method are analyzed. Theoretical analysis demonstrates that the method is faster than the FFT-convolution method for one time. The simulation experiment demonstrates that the method can advanced the filtering speed for about one time compared to the FFT-convolution method in the condition of no impacting the reconstructed images precision.
- Image reconstruction /
- Filtered Back Projection (FBP) /
- Acceleration /
- Hadamard Transform (HT) /
- Convolution

HTML全文

参考文献(1)

[1] 乔志伟, 魏学业, 韩焱. 解析法图像重建中的插值技术研究[J].计算机工程与设计, 2009, 30(9): 2213-2216. Qiao Zhi-wei, Wei Xue-ye and Han Yan. Study on interpolation technology in image reconstruction based on analytic method[J]. Computer Engineering and Design, 2009, 30(9): 2213-2216. [2] 种稚萌, 朱世华, 吕刚明. 多径信道下的异步分布式线性卷积空时编码[J].电子与信息学报.2009, 31(6):1415-1419浏览 Zhong Zhi-meng, Zhu Shi-hua, and L Gang-ming. Asynchronous distributed linear convolutional space-time code under multipath channels[J].Journal of Electronics Information Technology.2009, 31(6):1415-1419 [3] 虞湘宾, 毕光国. 长序列信号快速相关及卷积的算法研究[J].电路与系统学报, 2001, 6(4): 78-82. [4] Yu Xiang-bin and Bi Guang-guo. Algorithms of long sequence fast correlation and convolution[J]. Journal of Circuits and [5] Systems, 2001, 6(4): 78-82. [6] 吕新华, 武斌. 基于圆周卷积的长序列小波变换快速实现[J].信号处理, 2006, 26(2): 903-905. L Xin-hua and Wu Bin. Fast implementation of long sequence wavelet transform based on cyclic convolution[J]. Signal Proccessing, 2006, 26(2): 903-905. [7] 黄晓萍，桑恩方，乔钢. H序沃尔什变换及其在水声扩频通信中的应用[J].声学技术.2007, 26(3):477-482 Huang Xiao-ping, Sang En-fang, and Qiao Gang. Fast H-order Walsh transform and its applications in underwater acoustic spread-spectrum communication[J]. Technical Acoustics, 2007, 26(3): 477-482. [8] 李何明, 张大兴. 一种基于Hadamard变换的快速盲水印算法[J]. 杭州电子科技大学学报, 2009, 29(1): 67-70. Li He-ming and Zhang Da-xing. A blind fast image watermarking method based on Hadamard transform[J]. Journal of Hangzhou Dianzi University, 2009, 29(1): 67-70. [9] 张朝宗，郭志平，张朋. 工业CT技术和原理[M]. 北京: 科学出版社, 2009: 第8章. [10] Zhang Chao-zong, Guo Zhi-ping, and Zhang Peng. Industrial CT Technology and Principle[M]. Beijing: Science Press, 2009, Chapter 8. [11] 王召巴. 基于面阵CCD相机的高能X射线工业CT技术研究[D]. [博士论文], 南京理工大学, 2002. [12] Wang Zhao-ba. Study on high energy X-ray industrial CT technology based on area-CCD[D]. [Ph.D. dissertation], Nanjing University of Technology, 2002. [13] Zarowski C and Yunik M. Spectral filtering using the fast walsh transform[J].IEEE Transactions on Acoustics, Speech, and Signal Processing.1985, 33(4):1246-1252 [14] Thomas G and Govindan V K. Computationally efficient filtered-backprojection algorithm for tomographic image reconstruction using walsh transform[J].Journal of Visual Communication Image Representation.2006, 17(3):581-588

施引文献

资源附件(0)

访问统计