【摘要】 针对Active Demons算法只能配准同模态图像,并且图像的灰度梯度不太明显时形变方向无法确定从而导致误配准的问题,提出了一种基于局部熵改进的Active Demons多模医学图像配准算法.该算法首先将原图像转换为局部熵图像,然后使用改进的Active Demons算法进行配准,该算法在Active Demons分母上分别加入浮动图像与参考图像的灰度梯度值平方.正则化变换时以双边滤波器代替高斯滤波器,使得图像边缘的细节信息保持较好.多模态医学图像配准的实验结果说明该方法精确度高,使图像质量得以提升.更多还原

【Abstract】 In view of Active Demons algorithm is limited to the same mode image registration, and when the gray gradient information of the image is not obvious, the direction of image deformation can not be determined, an improved Active Demons multi-modality medical image registration based on local entropy image representation is proposed. Firstly, the image is converted into entropy image, and then the improved Active Demons algorithm, which the denominator is added with the square of the gray gradient of the floating image and the reference image, is used for registration. In the regularized transform, the paper use bilateral filter instead of Gauss filter to keep the detail edge information of the image better. The experimental results of multimodal medical image registration show that the proposed method is accurate and improves the image quality.更多还原