热流场的层析图是一些以条纹为主要特征的流场信息记录。条纹分析无论对干涉层析术还是对莫尔层析术都是极为关键的步骤。如何从各种噪声尤其是强烈的不均匀背景光噪声中提取清晰完整的条纹是热流场层析图预处理的主要问题。本文从热流场层析图的特征入手，修正灰度直方图重心，引入可提取概率作为区分噪声和信号的判据，将灰度图像依据由图像性质决定的投影函数投影到概率平面上，进行迭代以增强信号，衰减噪声。事实证明，本文所提出的方法对热流场层析图的预处理是行之有效的。 The chromatograms of the heat flow field are some of the flow field information records with the stripe as the main feature. Streak analysis is an extremely crucial step for either interferometry or Moiré surgery. How to extract clear and complete fringes from various kinds of noises, especially intense inhomogeneous background light noise, is the main problem in the preprocessing of heat flow field tomograms. In this paper, we start with the characteristics of the thermal flow field tomogram, modify the center of gravity of the grayscale histogram, introduce the extractable probability as the criterion to distinguish the noise and the signal, project the grayscale image onto the probability plane according to the projection function determined by the image property, Iterate to enhance the signal and attenuate the noise. It turns out that the method proposed in this paper is effective for the pretreatment of thermal flow field chromatogram.