Abstract: To improve the poor availability of elastic model in multimodality image registration，an image model translation method is presented.The mean value of a group of pixels which corresponds to the same pixel value in another image is calculated.Corresponding pixel value in the original image is replaced by the mean value.The gray of two images are converted into the same modal.This paper conducts image registration between the converted grayscale image and the target image，and realizes multimodality image registration by mapping the deformation parameter to the floating image.The experimental results show that the method can successfully use the local affine model for multimodal image registration.

Key words: elastic image registration, multimodality image, local affine model

摘要： 针对目前弹性图像配准方法较难应用于多模态图像的问题，提出了一种转化图像模态的解决方案。计算图像中每个灰度值在另一幅图像中对应像素的灰度均值，使用该均值代替原图像中对应的像素值，两幅图像灰度被转换为基本一致的状态，使用局部仿射模型配准图像。将灰度转换后的图像与目标图像配准，再将图像的形变参数映射到浮动图像中就可以实现多模态的图像配准。实验结果表明该方法可将局部仿射模型成功地用于多模态图像配准。

关键词: 弹性图像配准, 多模态图像, 局部仿射模型