基于贝叶斯准则的闪光照相图像重建  

Reconstruction of flash radiographic image based on Bayesian approach

作　　者：管永红[1] 王鹏来[1] 景越峰[1] 

机构地区：[1]中国工程物理研究院流体物理研究所,四川绵阳621900

出　　处：《强激光与粒子束》2011年第9期2495-2498,共4页High Power Laser and Particle Beams

基　　金：国家自然科学基金项目(61007065)

摘　　要：针对闪光照相图像低信噪比的特点,研究了一种带约束的贝叶斯图像重建算法,较好地抑制了噪声对重建结果的影响,同时对细节信号具有较高的保持能力。该算法以贝叶斯重建为基础,在重建过程中考虑了闪光照相图像的受模糊影响比较严重的特点,引入了重建结果具有平滑性的先验信息进行约束,以提高重建结果的信噪比。通过在迭代过程中对重建结果进行平滑性约束以提高重建质量。数值模拟结果表明,该方法在闪光照相图像重建中具有较好的抗噪能力和边界保持能力。A reconstruction algorithm is proposed for flash radiographic image with low signal-to-noise ratio.The algorithm is based on Bayesian principle,and uses smoothness constraint to handle the blurring effects caused by the X-ray source of radiography and imaging device.It is implemented with regularization schemes for performing tomographic image reconstruction from a single noisy projection.Numerical experimental results show that,in comparison with the conventional algebraic reconstruction technique,the algorithm has stronger ability of noise removal and edge preservation.

关 键 词：闪光照相 图像处理 图像重建 贝叶斯准则 

分 类 号：TN919[电子电信—通信与信息系统]