Abstract:Rigid robot pose estimation with an RGB-D camera has attracted substantial research attention recently for the reason that the RGB-D camera can capture RGB and depth information simultaneously. Despite the huge progress that has been made, there are still some unresolved issues like the pose estimation in texture-less or structure-less scenes. Aiming at this problem, this paper presents a robust real-time pose estimation method with an RGB-D camera for texture-less and structure-less scenes. Our contributions are threefold. First, we present an improved ORB algorithm for extracting reliable inliers, in which adaptive threshold setting method of FAST corners decision is proposed for extracting sufficient keypoints. In addition, an effective inliers refinement method, based on motion smoothness consistency constraint, is introduced for obtaining fine inliers. Second, based on the characteristics of RGB-D camera, this paper proposes a novel hybrid reprojection errors optimization model (HREOM) to estimate pose by concurrently minimizing 3D-3D and 3D-2D reprojection errors. Third, we carry out comprehensive experiments on TUM public datasets to demonstrate the robustness, accuracy, and real-time of the proposed system. The quantitative evaluations show that our system can extract sufficient inliers in those extreme scenes. Furthermore, our method performs as good as or better than other state-of-the-art solutions. Notably, our system can operate in the texture-less and structure-less environment, while other methods are prone to failure.

PoseRBPF: A Rao-Blackwellized Particle Filter For6d Object Pose Estimation.

PoseRBPF: A Rao-Blackwellized Particle Filter for 6D Object Pose Tracking

3D Point-to-Keypoint Voting Network for 6D Pose Estimation

Recurrent Volume-based 3D Feature Fusion for Real-time Multi-view Object Pose Estimation

Coupled 3D Tracking and Pose Optimization of Rigid Objects Using Particle Filter

Simultaneous robot Localization and Person Tracking using Rao-Blackwellised Particle Filters with multi-modal sensors

Physics-Based Object 6D-Pose Estimation during Non-Prehensile Manipulation

Rao-blackwellized particle filters for robot simultaneous localization and mapping

Novel Rao-Blackwellized Particle Filter for Mobile Robot SLAM Using Monocular Vision

FEIF: Feature Excitation and Interactive Fusion for 6D Object Pose Estimation.

PA-Pose: Partial Point Cloud Fusion Based on Reliable Alignment for 6D Pose Tracking

RFFCE: Residual Feature Fusion and Confidence Evaluation Network for 6dof Pose Estimation.

Robust FastSLAM Algorithm for Mobile Robot with an Adaptive Rao-Blackwellized Particle Filter

Gaussian Particle Filter Based Pose and Motion Estimation

Pose Estimation Based on a Dual Quaternion Feedback Particle Filter

6D Pose Estimation for Textureless Objects on RGB Frames using Multi-View Optimization

Research on the improvement of Rao-Blackwellized particle filter for the incremental environment mapping and self-localization of a mobile robot

Improved Rao-Blackwellized particle filters for mobile robot simultaneous localization and mapping

Visually Guided Model Predictive Robot Control via 6D Object Pose Localization and Tracking

Robust and Accurate Monocular Pose Tracking for Large Pose Shift.

Robust Robot Pose Estimation for Challenging Scenes With an RGB-D Camera