Realization of the Monocular Vision Odometry Based on Extracting Keypoints by SIFT in the Navigation System

A monocular vision odometry way based on the SIFT(Scale Invariant Feature Transform)used in the navigation system is designed.Deviations brought by the traditional odometry in the slippery surroundings can be choked back,and the course information can also be got.To realize the design,only one camera is needed.The scale invariant and the rotation invariant property of the SIFT are well used to extract the keypoints.By accurately matching the keypoints from the series images,estimations of the distance and rotation angles of the vehicle are got by the transformation of coordinates.The experiments show this way is simpler and the result is good.