基于三角剖分的视觉里程计特征点匹配算法
|
摘要
针对单目视觉里程计的精确定位问题,提出一种专门应用于单目视觉里程计的特征点匹配方法。对FAST算法提取的特征点进行DELAUNAY三角剖分,获得特征点的位置关系。采用LK算法获得匹配点,结合位置关系判断误匹配候选点。对候选点计算SIFT描述子,根据相似程度剔除误匹配点。对获得的匹配对估计基础矩阵,结合尺度信息求取位姿。实验结果表明:该算法可提升匹配点的正确率,提高单目视觉里程计的精度。
Aiming at the problem of precise positioning of monocular vision odometer,a method of feature point matching specifically applied to monocular vision odometer is proposed. Firstly,DELAUNAY triangulation is performed on the feature points extracted by the FAST algorithm to obtain the position relationship of the feature points. Afterwards,the matching points are obtained by using LK algorithm,and mis-matching candidate points are judged according to the position relation. Then,the SIFT descriptor is calculated for the candidate points and the mis-matching points are eliminated according to similarities. Finally,estimate basic matrixes for obtained matching pair,obtain the pose based on the scale information. Experimental results show that the algorithm can improve the accuracy of matching points and improve the precision of monocular vision odometer.
Aiming at the problem of precise positioning of monocular vision odometer,a method of feature point matching specifically applied to monocular vision odometer is proposed. Firstly,DELAUNAY triangulation is performed on the feature points extracted by the FAST algorithm to obtain the position relationship of the feature points. Afterwards,the matching points are obtained by using LK algorithm,and mis-matching candidate points are judged according to the position relation. Then,the SIFT descriptor is calculated for the candidate points and the mis-matching points are eliminated according to similarities. Finally,estimate basic matrixes for obtained matching pair,obtain the pose based on the scale information. Experimental results show that the algorithm can improve the accuracy of matching points and improve the precision of monocular vision odometer.
引文
[1] SCARAMUZZA D,FRAUNDORFER F. Visual odometry(Part I:The First 30 Years and Fundamentals)[J]. IEEE Robotics&Automation Magazine,2011,18(4):80-92.
[2] LUO J,GWUN O. A comparison of SIFT,PCA-SIFT and SURF[J].International Journal of Image Processing,2013,3(4):143-152.
[3] BARRON J L,FLEET D J,BEAUCHEMIN S S. Performance of optical flow techniques[C]∥Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition,CVPR’92,IEEE,1992:236-242.
[5]袁小翠,吴禄慎,陈华伟. Delaunay三角剖分算法改进与对比分析[J].计算机应用与软件,2016,33(9):163-166.
[4]刘欣.基于局部不变特征的图像匹配算法研究[D].哈尔滨:哈尔滨理工大学,2017.
[6] SUN C,VALLOTTON P. Fast linear feature detection using multiple directional non-maximum suppression[C]∥International Conference on Pattern Recognition,IEEE,2006:288-291.
[7] LOWE D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision,2004,60(2):91-110.
[8]张永祥,古佩强,穆铁英.改进的RANSAC基础矩阵估计算法[J].小型微型计算机系统,2016,37(9):2084-2087.
[9] ROUSSEEUW P J,LEROY A M. Robust regression&outlier detection[J]. Journal of the American Statistical Association,1987,31(2):260-261.
[10]罗桂娥,李映.基于均匀角点匹配的基础矩阵估计方法[J].微型机与应用,2013,32(7):8-9.
[11]马良荣,张德澄.矩阵SVD分解法在工程数值计算中的应用[J].宁夏大学学报:自然版,1998(2):125-127.
[12] REBAY S. Efficient unstructured mesh generation by means of Delaunay triangulation and Bowyer-Watson algorithm[J]. Journal of Computational Physics,1993,106:1.
[13] GEIGER A. Are we ready for autonomous driving? The KITTI vision benchmark suite[C]∥IEEE Conference on Computer Vision and Pattern Recognition,IEEE Computer Society,2012:3354-3361.
[2] LUO J,GWUN O. A comparison of SIFT,PCA-SIFT and SURF[J].International Journal of Image Processing,2013,3(4):143-152.
[3] BARRON J L,FLEET D J,BEAUCHEMIN S S. Performance of optical flow techniques[C]∥Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition,CVPR’92,IEEE,1992:236-242.
[5]袁小翠,吴禄慎,陈华伟. Delaunay三角剖分算法改进与对比分析[J].计算机应用与软件,2016,33(9):163-166.
[4]刘欣.基于局部不变特征的图像匹配算法研究[D].哈尔滨:哈尔滨理工大学,2017.
[6] SUN C,VALLOTTON P. Fast linear feature detection using multiple directional non-maximum suppression[C]∥International Conference on Pattern Recognition,IEEE,2006:288-291.
[7] LOWE D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision,2004,60(2):91-110.
[8]张永祥,古佩强,穆铁英.改进的RANSAC基础矩阵估计算法[J].小型微型计算机系统,2016,37(9):2084-2087.
[9] ROUSSEEUW P J,LEROY A M. Robust regression&outlier detection[J]. Journal of the American Statistical Association,1987,31(2):260-261.
[10]罗桂娥,李映.基于均匀角点匹配的基础矩阵估计方法[J].微型机与应用,2013,32(7):8-9.
[11]马良荣,张德澄.矩阵SVD分解法在工程数值计算中的应用[J].宁夏大学学报:自然版,1998(2):125-127.
[12] REBAY S. Efficient unstructured mesh generation by means of Delaunay triangulation and Bowyer-Watson algorithm[J]. Journal of Computational Physics,1993,106:1.
[13] GEIGER A. Are we ready for autonomous driving? The KITTI vision benchmark suite[C]∥IEEE Conference on Computer Vision and Pattern Recognition,IEEE Computer Society,2012:3354-3361.