IMU-based slam motion blur pose tracking algorithm

Abstract

The present invention discloses an IMU-based SLAM motion fuzzy pose tracking algorithm, which includes the following steps: S1, extracting ORB features from the input image sequence, judging the type of the image sequence according to the number of feature points, and selecting one to perform S2 according to the judgment result Or S3 step; S2, if it is a normal image, use the uniform motion model to estimate the initial pose of the camera, and perform bundle adjustment of the motion parameters; S3, if it is a motion blurred image, use the IMU motion equation to obtain the estimated pose, and use the extended Kalman The filter obtains the optimized pose, and the estimated pose is combined with the optimized pose to obtain the final pose of the camera. The present invention aims at the problem that SLAM cannot perform camera positioning and tracking loss on the motion blur segment in the image sequence, combines and utilizes the kinematic equation of the inertial measurement unit and the extended Kalman filter to calculate and optimize the camera pose, so that SLAM can obtain continuous and reliable cameras Pose localization and tracking.

Description

technical field [0001] The invention relates to a tracking algorithm, in particular to an IMU-based SLAM motion fuzzy pose tracking algorithm, belonging to the fields of computer vision and robots. Background technique [0002] Simultaneous localization and mapping (SLAM) is a research hotspot in the field of computer vision and robotics in recent years. SLAM technology can construct and update maps in unknown environments, and track and locate in real time. Early SLAM methods used filtering to solve problems. Davinson et al. proposed a real-time single-camera MonoSLAM method, which uses extended Kalman filtering as the backend to track very sparse feature points at the front end. Eade et al. proposed a scale-variable monocular SLAM method, which uses particle filtering and top-down search to draw a large number of landmarks in real time. Most of these methods use filters to process image frames to correlate and estimate the positions of map points and camera poses. Since ...

AI Technical Summary

Problems solved by technology

[0005] To sum up, how to provide an IMU-based SLAM motion blur pose tracking algorithm to solve the problem that SLAM cannot perform camera positioning and tracking loss on the motion blur segment in the image sequence has become the common expectation of technicians in the industry. one of the problems solved

Images