摘要
在使用CT进行图像重建的过程中,需要在不同角度下对目标对象进行采样,然后利用图像重建算法生成重建结果,由于采样的数据越多,重建速率越慢,往往需要在不完全的采样角度下对图像进行重建,即稀疏重建。为了对传统稀疏重建算法的迭代速度进行改进,在传统bregman图像重建算法的基础上提出了一种新的加速迭代算法。该算法以bregman算法为框架,结合自适应梯度下降算法和图像修正算法,从而实现了稀疏角度下的快速重建。实验结果表明,新的加速算法对在成像效果上具有比较好的结果,且收敛速度明显加快。
In the process of CT image reconstruction,we need to sample the target in different angles, and then rebuild the results generated by image reconstruction algorithm, as a result of the more data sampled, the slower the reconstruction rate, we often require image under incomplete sampling angle for image reconstruction. In order to improve the iteration speed of the traditional sparse reconstruction algorithm, based on the traditional bregman image reconstruction algorithm, we propose a new accelerated iterative algorithm. The algorithm is based on bregman algorithm, by combining the adaptive gradient descent algorithm and image correction algorithm, the fast reconstruction under sparse angles is realized. The experimental results show that the new algorithm has good results in the imaging effect, and the convergence speed is obviously accelerated.
作者
李欣
Li Xin(shanxi international business vocational college,Tai yuan 030031,China)
出处
《长江信息通信》
2021年第1期37-40,共4页
Changjiang Information & Communications
