论文部分内容阅读
太赫兹(THz)波具有的许多独特性质,使其非常适合应用于对人体的安检成像,但是目前原始太赫兹图像的信噪比、对比度和分辨率都有待改善。为提高太赫兹安检图像的质量,研究提出一种基于马尔可夫随机场理论的被动式太赫兹图像复原算法。对原始图像进行去噪和增强的预处理之后,在贝叶斯分析的基础上增加马尔可夫约束项进行图像复原。通过改变迭代次数和正则化参数,得到了清晰度不同的处理结果,经主客观评价指标分析确定了最佳的参数。实验结果证明,本算法可以在被动式太赫兹图像的噪声滤除和边缘信息保持上取得较好的平衡,从而大幅提高太赫兹安检图像的目标分辨能力。
THz waves have many unique properties that make them ideal for security imaging of the human body, but the current terahertz images of the signal to noise ratio, contrast and resolution have yet to be improved. In order to improve the quality of terahertz security image, a passive terahertz image restoration algorithm based on Markov random field theory is proposed. After the original image is denoised and enhanced, the Markov constraint is added to the image restoration based on the Bayesian analysis. By changing the number of iterations and regularization parameters, the results of different resolutions are obtained, and the optimal parameters are determined through the analysis of subjective and objective evaluation indexes. Experimental results show that the proposed algorithm can achieve a good balance between noise filtering and edge information preservation in passive terahertz images, which can greatly improve the target resolution of terahertz security images.