INS/vSLAM System Implementation using a Feature Point Matching Filter-Combined SIFT

S. Yun, Y.J. Lee, S. Sung, and M. Gwak (Korea)


Relative position, INS/vSLAM, integration filter, IMU,Vision sensor, navigation


This paper aims to implement a target tracking system using a feature point matching filter-combined SIFT algorithm and analyze tracking performance of the system. The system is capable to operate in real-time manner. The presented algorithm is based on the SIFT algorithm and feature point matching filter which plays a role in eliminating outliers is newly applied. And multi-target tracking system is constructed using this algorithm based on LabVIEW programming in PC environment. Finally, INS/vSLAM integrated navigation system is constructed using the proposed feature point matching filter-combined SIFT algorithm and performance of the system is verified via indoor test by comparison between KLT algorithm and the proposed SIFT algorithm used as the feature point tracking method.

Important Links:

Go Back