通过改变入射的Ｘ射线能谱， 从而得到物体内部不同密度成分的分布图像的方法，是Ｘ辐射成像领域的一个重要课题。本文提出了一种新的图像分离方法， 即通过影像灰度进行密度分布图像重建， 从本质上对这种算法进行了探讨， 证明了这种算法的可行性， 并且进行了两种成分的图像分离实验，取得了较好的结果。这种方法的突出特点是减少了运算量和运算时间，可以快速实现图像分离， 有一定的实用性。 It is an important issue in the field of X-ray imaging by changing the incident X-ray energy spectrum to obtain the distribution images of different density components inside the object. In this paper, a new method of image separation is proposed, that is, density distribution image reconstruction by image gray level. The algorithm is discussed in essence, and the feasibility of this algorithm is proved. And two components images Separation experiment, and achieved good results. The salient features of this method are the reduction of computation and computation time, the rapid separation of images, and a certain degree of practicability.