摘要
通过分析H arris特征点检测算法在应用中的不足,提出一种基于图像分块的多尺度H arris特征点检测算法,解决了特征点聚簇现象,检测出的特征点分布均匀.实验结果表明,该算法具有精确性、有效性和鲁棒性,为进一步图像特征点匹配工作提供了保证.
Based on the shortcoming of Harris algorithm in application, a multi-scale Harris algorithm for the feature point detection based on image sub-block is proposed. It makes the detected feature points well-proportioned distributed without clusters. Experiments demonstrate that this algorithm has a good performance of accuracy,validity and robustness, which provides a guarantee for the further matching of feature points
出处
《内蒙古大学学报（自然科学版）》
CAS
CSCD
北大核心
2009年第3期326-329,共4页
Journal of Inner Mongolia University：Natural Science Edition
基金
教育部"春晖计划"资助项目(Z2007-1-01010)
