论文部分内容阅读
本文引用了一种图像分布的先验模型，在这种先验模型中引入了两个边界模型，称为组合马尔柯夫随机场模型，用Ｂａｙｅｓ定理将其与放射ＣＴ（ＥＣＴ）问题相结合构造出后验概率，由于这个最大后验估计问题是一个非凸问题，本文采用平均场退火技术对其进行分析，将统计松弛方程转化为一组确定性迭代式，计算机进行的模拟实验证实了这种算法的有效性。
This paper cites a priori model of image distribution. In this prior model, two boundary models are introduced, which are called combinatorial Markov random field model, which is combined with radiological CT (ECT) problem by Bayes theorem Since posterior probability is a non-convex problem, mean-field annealing technique is used in this paper to analyze the statistical relaxation equation into a set of deterministic iterative methods. Computer simulations confirm The validity of this algorithm.