本文采用基于特征点的匹配算法完成对实验医学图像的匹配，从特征点数量、特征提取时间和匹配准确性等方面比较尺度不变特征变换（SIFT）和快速鲁棒特征（SURF）算法，然后采用K最近邻算法（KNN）去除误匹配，统计分析不同阈值的随机抽样一致算法（RANSAC）和最小中值方差估计算法（LMEDS）与配准结果的相关性。本研究建立了基于特征点的医学图像配准算法程序实验平台，实现了多算法融合的医学图像特征匹配，对进一步探讨和改进医学图像配准提供了研究基础。%The paper adopted the matching algorithm based on characteristic points to accomplish image matching in experimental medicine. The two algorithms: scale-invariant feature transform (SIFT) and speeded up robust features (SURF) were compared in aspects of characteristic points, characteristic extraction time and matching veracity. Then the K-nearest neighbor (KNN) algorithm was used to eliminate mismatching points. The paper also carried on statistical analysis of the results of random sample consensus (RANSAC) and least median of squres (LMEDS) as well as the correlation between matching veracity and algorithms. This research established a medical image matching platform based on characteristic points in order to provide a research basis for the further research and improvement of medical image matching.
展开▼
