VDO-SLAM: A Visual Dynamic Object-aware SLAM System

Abstract

The scene rigidity assumption, also known as the static world assumption, is common in SLAM algorithms. Most existing algorithms operating in complex dynamic environments simplify the problem by removing moving objects from consideration or tracking them separately. Such strong assumptions limit the deployment of autonomous mobile robotic systems in a wide range of important real world applications involving highly dynamic and unstructured environments. This paper presents VDO-SLAM, a robust object-aware dynamic SLAM system that exploits semantic information to enable motion estimation of rigid objects in the scene without any prior knowledge of the objects shape or motion models. The proposed approach integrates dynamic and static structures in the environment into a unified estimation framework resulting in accurate robot pose and spatio-temporal map estimation. We provide a way to extract velocity estimates from object pose change of moving objects in the scene providing an important functionality for navigation in complex dynamic environments. We demonstrate the performance of the proposed system on a number of real indoor and outdoor datasets. Results show consistent and substantial improvements over state-of-the-art algorithms. An open-source version of the source code is available * .

AirDOS: Dynamic SLAM benefits from Articulated Objects

“…As shown in the Table II, with human masks, our algorithm obtains a 39.5% and 15.2% improvements compared to ORB-SLAM [3] and VDO-SLAM [30] in the overall performance. In Sequence II, IV and V, both ORB-SLAM and our algorithm show a good performance, where all ATEs are lower than 0.04.…”

“…We use the Absolute Translation Error (ATE) to evaluate our algorithm. Our method is compared against the state-ofthe-art methods, ORB-SLAM [3] (1) with and ( 2) without the masking of potential dynamic objects, and RGB-D dynamic SLAM algorithm [30] which inherits from [7], [9]. Our algorithm is implemented based on ORB-SLAM [3].…”

Jacobian Computation for Cumulative B-Splines on SE(3) and Application to Continuous-Time Object Tracking

“…For DT, as it is common in the literature [8], [10], we assume constant kinematics between two time stamps, with t k+1 = t k + ∆t. The linear (v o ) and angular (ω o ) velocity at t ∈ [t k , t k+1 ), for coupled and decoupled translation and rotation, are given by Eqs.…”

“…With these goals in mind, we find in the literature several approaches that aim to estimate the position and orientation of the objects (SE(3) poses) [7], [8], [9]. However, relevant kinematic magnitudes such as the linear and angular velocities are not considered in the estimated model.…”

Autonomous Vehicular Networks: Perspective and Open Issues

“…4 shows a pipeline of the performance evaluation system that combines the sensing and inter-vehicle communications, which showcases the positive effect of sensing on V2V communications. The sensor data processing components are achieved by the basic SLAM system with object detection capability [12]. The detected objects and extracted object features are shown in Fig.…”