基于局部锐度分布耦合核典型相关分析的图像匹配算法
On Image Matching Algorithm Based on Local Sharpness Distribution and Kernel Canonical Correlation Analysis Method
-
摘要: 提出了基于局部锐度分布耦合核典型相关分析的图像匹配算法.首先引入Forstner算子对特征点进行精确提取;随后计算每个特征点对应的锐度值,从而构造局部锐度分布模型,生成低维度的特征描述子;接着引入归一化互相匹配策略(Normalized Cross Correlation,NCC),完成特征点的匹配,增强算法的鲁棒性;最后基于核典型相关分析(Kernel Canonical Correlation Analysis,KCCA)技术,建立归一化距离函数,对匹配特征点进行提纯,剔除误匹配点.仿真实验结果表明:与当前图像匹配算法相比,本文算法不仅具有较高的匹配精度及较强的鲁棒性,而且还具有较高的匹配效率.
-
关键词:
- 图像匹配 /
- 局部锐度分布 /
- 核典型相关分析 /
- Forstner算子 /
- 归一化距离函数
Abstract: The image matching algorithm based on local sharpness distribution and kernel canonical correlation analysis method has been proposed in this paper.Firstly,the feature points were accurately extracted by introducing the Forstner operator.Then the local sharpness distribution model was constructed by calculating each feature point sharpness value to generate low dimension feature descriptor.the matching of feature points were finished by introducing the normalized matching strategy to enhance the robustness.And finally,the normalized distance function was established based on the kernel canonical correlation analysis purifies the matching feature points to eliminate the false matching points.The simulation results show that: this algorithm not only has high matching accuracy and strong robustness,but also has a high matching efficiency compared with the current image matching algorithm. -
-
计量
- 文章访问数: 825
- HTML全文浏览数: 456
- PDF下载数: 69
- 施引文献: 0
下载: