中文摘要：

数字减影图像矫正是DSA血管三维重建的重要预处理过程.运用分阶段矫正思想,本文首次提出基于加权直方图的数字减影图像矫正算法,对传统运动矫正和灰度矫正算法均进行有效改进.改进措施包括根据减影图像边缘信息选取控制点、构造基于减影加权直方图能量的相似性测度、生成血管排除模板和利用减影加权直方图寻找灰度偏移量.实验结果表明,改进算法能改善减影图像质量,增强血管辨识度,更有效消除运动噪声和灰度噪声,矫正效果明显.

英文摘要：

Image registration for digital subtraction angiography （DSA） is an important preprocess for 3D reconstruction of blood vessels based on DSA image sequence. A new image registration algorithm based on weighted histogram for DSA image is firstly proposed in this paper. The registration method is divided into two phases,i, e. motion correction and gray-level registration, according to the two different types of noise. A series of methods are introduced to improve the performances of traditional DSA registration algorithms, including selection of control points on the basis of edges in subtraction image, construction of new similarity measure based on the energy of weighted histogram of differences, generation of blood exclusion template and determination of gray value shifts according to the new weighted histogram of subtraction image. Various experimental results have demonstrated that the proposed algorithm is able to eliminate the motion noises and gray noises more effectively. The visualization of blood vessels is enhanced and the quality of subtraction image registration is improved therefore.