提出结合分段对比度拉伸变换(PCST)和基于块的局部最优维纳(PLOW)滤波算法进行太赫兹Gabor同轴数字全息二维再现像图像复原的方法。分别介绍了PCST和PLOW的工作原理， 给出了PCST-PLOW组合滤波算法的工作流程。对真实的2.52 THz数字全息再现像进行了去噪处理，并对每一处理过程进行了对比分析。比较了PCST节点对增强效果的影响,其节点的起点和终点分别根据直方图和图像均值确定。对真实太赫兹再现像的处理结果显示：去噪处理后的40帧平均全息图再现像的整体标准差由0.184 9提升到0.307 4，均值与标准差的比值由1.720 8提升到2.268 8。得到的结果表明所用方法明显抑制了噪声，同时提高了图像对比度。

An image restoration method of 2D reconstructed-image of terahertz Gabor in-line digital holography was proposed with combination of Piecewise-contrast-stretching Transformation(PCST) and Patch-based Locally Optimal Wiener (PLOW) filter. The PCST and the PLOW algorithms were introduced, respectively, and the working flow of the combined PCST-PLOW algorithm was given. The denoising for a real 2.52 THz reconstructed-image of digital holography was executed, and each step of the image processing was compared. The effect of the PCST on the image enhancement was analysis, in which the starting point and the end point of the nodes were dependent on a histogram and an image average value. The results of treatment on the real terahertz reconstructed-image show that the global standard deviation of the denoising reconstructed-image of 40-frame averaged hologram rises from 0.184 9 to 0.307 4, and the ratio of the mean value to standard deviation rises from 1.720 8 to 2.268 8. These results demonstrate that the proposed method has suppressed image noise, while improving image contrast.