基于中心环绕滤波器检测的图像特征点匹配算法

发布时间：2018-07-22 15:25

【摘要】：针对传统图像匹配算法特征点检测稳定性和准确性差的问题,提出一种尺度不变性的基于中心环绕滤波器检测(SCFD)的图像特征点匹配算法。首先,构建多尺度空间,利用中心环绕滤波器检测图像在不同尺度下的特征点,采用Harris方法和亚像素插值获得稳定的特征点;其次,联合快速定向旋转二进制稳健基元独立特征(BRIEF)(ORB)算法确定特征点的主方向,构建特征点描述算子;最后,采用汉明距离完成匹配,通过最小平方中值(LMed S)定理和最大似然(ML)估计剔除误匹配点。实验结果表明,在尺度变化时,所提算法的匹配精度达到96.6%,是ORB算法的2倍;其运行时间是尺度不变特征变换(SIFT)的19.8%,加速鲁棒性特征(SURF)的28.3%。所提算法能够有效提高特征点检测的稳定性和准确性,在视角、尺度缩放、旋转、亮度等变化的情况下具有较好的匹配效果。
[Abstract]:A scale-invariant image feature point matching algorithm based on center surround filter (SCFD) is proposed to solve the problem of poor stability and accuracy of the traditional image matching algorithm. Firstly, the multi-scale space is constructed, and the feature points of the image at different scales are detected by using the center surround filter, and the stable feature points are obtained by Harris method and sub-pixel interpolation. Combined with the fast directed rotation binary robust primitive independent feature (Orb) algorithm to determine the main direction of feature points and construct feature point description operator. Finally, the hamming distance is used to complete the matching. The mismatch points are eliminated by the least square mean value (LMed S) theorem and the maximum likelihood (ML) estimation. The experimental results show that the matching accuracy of the proposed algorithm is 96. 6, which is twice that of the Orb algorithm, and its running time is 19. 8% of the scale invariant feature transform (sift), and the accelerated robustness feature (SURF) is 28. 3% of the scale invariant feature transform (sift). The proposed algorithm can effectively improve the stability and accuracy of feature point detection, and has a better matching effect in the case of angle of view, scale scaling, rotation, brightness and so on.
【作者单位】：东北电力大学信息工程学院;
【分类号】：TP391.41

【相似文献】