Feature-based SLAM for Dense Mapping

An improved feature-based SLAM for dense mapping is proposed in this paper to easy follow-up works of robots. Features are extracted by original ORB-SLAM for pose estimation. Filter is used to get gradient graph for each image and points whose gradient are higher than threshold is added as feature points, then a voxel filter is adopted to reduce sampling. Maps are built by method as in ORB-SLAM but denser. Experimental results demonstrate the improved feature-based SLAM in this paper reconstructs scenes well with dense maps with acceptable CPU usage and efficiency.

[1]  Hujun Bao,et al.  ICE-BA: Incremental, Consistent and Efficient Bundle Adjustment for Visual-Inertial SLAM , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[2]  Javier Civera,et al.  Inverse Depth Parametrization for Monocular SLAM , 2008, IEEE Transactions on Robotics.

[3]  Michael Gassner,et al.  SVO: Semidirect Visual Odometry for Monocular and Multicamera Systems , 2017, IEEE Transactions on Robotics.

[4]  Daniel Cremers,et al.  LSD-SLAM: Large-Scale Direct Monocular SLAM , 2014, ECCV.

[5]  Olivier Stasse,et al.  MonoSLAM: Real-Time Single Camera SLAM , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Juan D. Tardós,et al.  Visual-Inertial Monocular SLAM With Map Reuse , 2016, IEEE Robotics and Automation Letters.

[7]  King-Sun Fu,et al.  IEEE Transactions on Pattern Analysis and Machine Intelligence Publication Information , 2004, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Shaojie Shen,et al.  Monocular Visual-Inertial State Estimation for Mobile Augmented Reality , 2017, 2017 IEEE International Symposium on Mixed and Augmented Reality (ISMAR).

[9]  J. M. M. Montiel,et al.  ORB-SLAM: A Versatile and Accurate Monocular SLAM System , 2015, IEEE Transactions on Robotics.