摘要
提出了一种基于四元数的彩色图像匹配算法.该算法将彩色图像用四元数矩阵模型描述,抽取四元数矩阵的奇异值作为匹配特征.对奇异值进行去噪处理,增+了匹配算法的抗噪能力.采用分层搜索策略,减少运算量,提高了匹配速度.实验结果验证了方法的有效性.
A quaternion-based color image matching algorithm has been proposed.Color image is described by quaternion matrices model in this algorithm,and the singular value of the quaternion matrices is extracted as matching feature.De-noise processing to the singular value enhances noise immunity.Hierarchical search strategy is used to reduce computation and improve the matching speed.Experimental results demonstrate the effectiveness of this method.
出处
《嘉应学院学报》
2011年第2期30-32,共3页
Journal of Jiaying University
关键词
彩色图像匹配
四元数矩阵
奇异值分解
分层搜索
color image matching
quaternion matrix
singular value decomposition
hierarchical search
