基于FAST-SURF的移动端实时特征检测匹配算法

Real-time feature detecting and matching algorithm based on FAST-SURF on mobile terminal

针对移动终端自然特征提取和匹配的处理能力不足,提出了一种基于FAST-SURF特征检测匹配的简化算法,采用FAST算法检测自然特征点,再采用简化过的SURF算法计算特征点的方向,并建立特征描述符,然后将建立好的特征点描述符与数据库中的特征描述符进行匹配。实验结果表明,该算法处理图像花费时间短,在移动端能实时处理自然特征检测与匹配。

In this paper,a simplified algorithm based on FAST-SURF is proposed in order to solve the problem of low processing capacity of mobile terminal on natural feature detection and matching. The FAST algorithm is used to detect natural features. The simplified SURF algorithm is used to calculate the orientation and describe the feature points. Then the established feature descriptor is matched in the database. The experimental results show that this algorithm is fast enough to satisfy the requirement of real-time natural feature detection and matching on mobile terminal.