| 一种新的PET序列图像超分辨率优质重建算法 |
| |
| 引用本文： | 刘建琳,潘燕.一种新的PET序列图像超分辨率优质重建算法[J].实验室研究与探索,2008,27(8). |
| |
| 作者姓名： | 刘建琳 潘燕 |
| |
| 作者单位： | 山东大学医学院生物医学工程研究所,山东,济南,250012 |
| |
| 摘 要： | 利用超分辨率重建技术,从含有加性高斯噪声和模糊噪声的正电子发射成像(PET)序列低分辨率图像,重建出一幅优质高分辨率图像。作者提出了一种基于正则化参数(RP)的通道自适应线性斜率超分辨率算法。该算法采用平移运动模型,通过对RP线性斜率的自适应更新,动态优化代价函数,以降低对PET图像高频成分的抑制。为验证新算法的有效性,采用模拟PET序列图像进行实验。实验中,与HUHE算法相比,新算法PSNR平均提高2.65dB。新算法在改善图像空间分辨率上取得良好的效果,同时具有很好的抗噪性能。
|
| 关 键 词： | 自适应 线性斜率 超分辨率重建 PET图像 |
A Novel Super-Resolution High Quality Reconstruction Algorithm for PET Image Sequence |
| |
| LIU Jian-lin,PAN Yan.A Novel Super-Resolution High Quality Reconstruction Algorithm for PET Image Sequence[J].Laboratory Research and Exploration,2008,27(8). |
| |
| Authors: | LIU Jian-lin PAN Yan |
| |
| Abstract: | Super-resolution(SR) technique was used to estimate high-resolution(HR) images from a sequence of low-resolution(LR) positron emission tomography(PET) images that include additive Gaussian noise and blurred noise. A channel-adaptive linear slope algorithm based on the regularization parameter(RP) was proposed. This algorithm, using a translational motion model, can adaptively optimize the cost function and reduce the constraint high frequency components of PET image by adaptively updating linear slope of the RP. The convergence property of proposed algorithm was analyzed. The proposed method was performed on simulated PET image sequence. Compared to the method of HUHE, the average improvement for the proposed method is 2.65 dB. The proposed algorithm outperforms the conventional approach in the spatial resolution; meanwhile the anti-noise performance is improved in this novel method. |
| |
| Keywords: | adaptive linear slope super-resolution reconstruction PET image |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
