Full 6-DOF Research Articles

Wing Kinematics-Based Flight Control Strategy in Insect-Inspired Flight Systems: Deep Reinforcement Learning Gives Solutions and Inspires Controller Design in Flapping MAVs.

Flying insects exhibit outperforming stability and control via continuous wing flapping even under severe disturbances in various conditions of wind gust and turbulence. While conventional linear proportional derivative (PD)-based controllers are widely employed in insect-inspired flight systems, they usually fail to deal with large perturbation conditions in terms of the 6-DoF nonlinear control strategy. Here we propose a novel wing kinematics-based controller, which is optimized based on deep reinforcement learning (DRL) to stabilize bumblebee hovering under large perturbations. A high-fidelity Open AI Gym environment is established through coupling a CFD data-driven aerodynamic model and a 6-DoF flight dynamic model. The control policy with an action space of 4 is optimized using the off-policy Soft Actor-Critic (SAC) algorithm with automating entropy adjustment, which is verified to be of feasibility and robustness to achieve fast stabilization of the bumblebee hovering flight under full 6-DoF large disturbances. The 6-DoF wing kinematics-based DRL control strategy may provide an efficient autonomous controller design for bioinspired flapping-wing micro air vehicles.

PrintShear

Most touch-based input devices, such as touchscreens and touchpads, capture low-resolution capacitive images when a finger touches the device's surface. These devices only output the two-dimensional (2D) positions of contacting points, which are insufficient for complex control tasks, such as the manipulation of 3D objects. To expand the modalities of touch inputs, researchers have proposed a variety of techniques, including finger poses, chording gestures, touch pressure, etc. With the rapid development of fingerprint sensing technology, especially under-screen fingerprint sensors, it has become possible to generate input commands to control multiple degrees of freedom (DOF) at a time using fingerprint images. In this paper, we propose PrintShear, a shear input technique based on fingerprint deformation. Lateral, longitudinal and rotational deformations are extracted from fingerprint images and mapped to 3DOF control commands. Further DOF expansion can be achieved through recognition of the contact region of the touching finger. We conducted a 12-person user study to evaluate the performance of PrintShear on 3D docking tasks. Comparisons with other input methods demonstrated the superiority of our approach. Specifically, a 19.79% reduction in completion time was achieved compared with conventional touch input in a full 6DOF 3D object manipulation task.

BoW3D: Bag of Words for Real-Time Loop Closing in 3D LiDAR SLAM

Loop closing is a fundamental part of simultaneous localization and mapping (SLAM) for autonomous mobile systems. In the field of visual SLAM, bag of words (BoW) has achieved great success in loop closure. The BoW features for loop searching can also be used in the subsequent 6-DoF loop correction. However, for 3D LiDAR SLAM, the state-of-the-art methods may fail to effectively recognize the loop in real time, and usually cannot correct the full 6-DoF loop pose. To address this limitation, we present a novel Bag of Words for real-time loop closing in 3D LiDAR SLAM, called BoW3D. Our method not only efficiently recognizes the revisited loop places, but also corrects the full 6-DoF loop pose in real time. BoW3D builds the bag of words based on the 3D LiDAR feature LinK3D, which is efficient, pose-invariant and can be used for accurate point-to-point matching. We furthermore embed our proposed method into 3D LiDAR odometry system to evaluate loop closing performance. We test our method on public dataset, and compare it against other state-of-the-art algorithms. BoW3D shows better performance in terms of F1 max and extended precision scores on most scenarios. It is noticeable that BoW3D takes an average of 48 ms to recognize and correct the loops on KITTI 00 (includes 4K+ 64-ray LiDAR scans), when executed on a notebook with an Intel Core i7 @2.2 GHz processor. We release the implementation of our method here: https://github.com/YungeCui/BoW3D.

A simplified method to consider hydrodynamic effect in oblique vessel-bridge collisions

The hydrodynamic effect in vessel-bridge collisions is either directly omitted or considered by the constant added mass (CAM) method in current studies. However, the high-resolution fluid-structure interaction (FSI) simulations indicate that the hydrodynamic effect plays an important role and cannot be adequately captured by the CAM method, particularly for oblique vessel-bridge collisions. On the other hand, the FSI simulations are time-consuming and have a high requirement for computational resources. For these reasons, a simplified analysis method was developed in this study to consider the hydrodynamic effect of vessel collision with bridge structures. In the developed method, mechanics of full 6-DoF ship maneuvering were established using several matrices to replace the explicit fluid domain of the FSI models, while the added mass method was adopted for the impacted pier to account for the hydrodynamic effect. Several collision scenarios were performed to compare the developed method with the FSI simulation and the CAM analysis. It was found that the proposed method is superior to the CAM method in terms of considering the hydrodynamic effect of vessel-bridge collisions, and its required computational cost can be significantly saved in comparison with the FSI model. From the perspective of calculation efficiency and accuracy, the proposed method is more suitable for the analysis and design of vessel-bridge collision in engineering practice than the FSI simulation and the CAM analysis.

Compliance modeling of a full 6-DOF series–parallel flexure-based Stewart platform-like micromanipulator

AbstractWith many micromanipulator designs emerging in micro and nanosystem applications, the element of compliance in the mechanisms is gaining attention. Several designs consider motions limited in a plane for high accuracy and repeatability as needed in micro/nano manipulation applications. Extending this to a full spatial configuration with coupled motions of series and parallel linkages with flexure joints of 1-degree-of-freedom (DOF) and 3-DOF needs a systematic analytical approach. One such approach for compliance analysis is presented in this article for a mechanism designed at Indian Institute of Technology Kharagpur. To validate the analytical models, finite element analysis simulations are performed with the help of the Abaqus-6.14 software package. Following the successful validation, the effect of structural parameters on the performance is presented with the help of the analytical expressions. We explore the performance of the mechanism with different dimensions of flexures of a particular type. Results indicate that the design with dissimilar dimensional parameters can give superior performance.

Learning-Based 6-DOF Control for Autonomous Proximity Operations Under Motion Constraints

This article proposes areinforcement learning (RL)-based six-degree-of-freedom (6-DOF) control scheme for the final-phase proximity operations of spacecraft. The main novelty of the proposed method are from two aspects: 1) The closed-loop performance can be improved in real-time through the RL technique, achieving an online approximate optimal control subject to the full 6-DOF nonlinear dynamics of spacecraft; 2) nontrivial motion constraints of proximity operations are considered and strictly obeyed during the whole control process. As a stepping stone, the dual-quaternion formalism is employed to characterize the 6-DOF dynamics model and motion constraints. Then, an RL-based control scheme is developed under the dual-quaternion algebraic framework to approximate the optimal control solution subject to a cost function and a Hamilton–Jacobi–Bellman equation. In addition, a specially designed barrier function is embedded in the reward function to avoid motion constraint violations. The Lyapunov-based stability analysis guarantees the ultimate boundedness of state errors and the weight of NN estimation errors. Besides, we also show that a PD-like controller under dual-quaternion formulation can be employed as the initial control policy to trigger the online learning process. The boundedness of it is proved by a special Lyapunov strictification method. Simulation results of prototypical spacecraft missions with proximity operations are provided to illustrate the effectiveness of the proposed method.

Is Halcyon feasible for single thoracic or lumbar vertebral segment SBRT?

PurposeHalcyon linear accelerators employ intensity‐modulated radiation therapy (IMRT) and stereotactic body radiation therapy (SBRT) techniques. The Halcyon offers translational, but not rotational, couch correction, which only allows a 3 degrees of freedom (3‐DOF) correction. In contrast, the TrueBeam (TB) linear accelerator offers full 6‐DOF corrections. This study aims to evaluate the difference in treatment plan quality for single thoracic or lumbar vertebral segment SBRT between the Halcyon and TB linear accelerators. In addition, this study will also investigate the effect of patient rotational setup errors on the final plan quality.MethodsWe analyzed 20 patients with a single‐level spine metastasis located between the T7 and L5 vertebrae near the spinal canal. The median planning target volume was 52.0 cm3 (17.9–138.7 cm3). The median tumor diameter in the axial plane was 4.6 cm (range 1.7–6.8 cm), in the sagittal plane was 3.3 cm (range 2–5 cm). The prescription doses were either 12–16 Gy in 1 fraction or 18–24 Gy in 3 fractions. All patients were treated on the TB linear accelerator with a 2.5 mm Multi‐Leaf Collimator (MLC) leaf width. Treatment plans were retrospectively created for the Halcyon, which has a 5 mm effective MLC leaf width. The 20 patients had a total of 50 treatments. Analysis of the 50 cone beam computed tomography (CBCT) scans showed average rotational setup errors of 0.6°, 1.2°, and 0.8° in pitch, yaw, and roll, respectively. Rotational error in roll was not considered in this study, as the original TB plans used a coplanar volumetric modulated arc therapy (VMAT) technique, and each 1° of roll will contribute an error of 1/360. If a plan has 3 arcs, the contribution from errors in roll will be < 0.1%. To simulate different patient setup errors, for each patient, 12 CT image datasets were generated in Velocity AI with different rotational combinations at a pitch and yaw of 1°, 2°, and 3°, respectively. We recalculated both the TB and Halcyon plans on these rotated images. The dosimetric plan quality was evaluated based on the percent tumor coverage, the Conformity Index (CI), Gradient Index (GI), Homogeneity index (HI), the maximum dose to the cord/cauda, and the volume of the cord/cauda receiving 8, 10, and 12 Gy (V8Gy, V10Gy and V12Gy). Paired t‐tests were performed between the original and rotated plans with a significance level of 0.05.ResultsThe Eclipse based VMAT plans on Halcyon achieved a similar target coverage (92.3 ± 3.0% vs. 92.4 ± 3.3%, p = 0.82) and CI (1.0 ± 0.1 vs. 1.1 ± 0.2, p = 0.12) compared to the TB plans. The Gradient index of Halcyon is higher (3.96 ±0.8) than TB (3.85 ±0.7), but not statistically significant. The maximum dose to the spinal cord/cauda was comparable (11.1 ± 2.8 Gy vs. 11.4 ± 3.6 Gy, p = 0.39), as were the V8Gy, V10Gy and V12Gy to the cord/cauda. The dosimetric influence of patient rotational setup error was statistically insignificant for rotations of up to 1° pitch/yaw (with similar target coverage, CI, max cord/cauda dose and V8Gy, V10Gy, V12Gy for cord/cauda). The total number of monitor units (MUs) for Halcyon (4998 ± 1688) was comparable to that of TB (5463 ± 2155) (p = 0.09).ConclusionsThe Halcyon VMAT plans for a single thoracic or lumbar spine metastasis were dosimetrically comparable to the TB plans. Patient rotation within 1° in the pitch and yaw directions, if corrected by translation, resulted in insignificant dosimetric effects. The Halcyon linear accelerator is an acceptable alternative to TB for the treatment of single thoracic or lumbar spinal level metastasis, but users need to be cautious about the patient rotational setup error. It is advisable to select patients appropriately, including only those with the thoracic or lumbar spine involvement and keeping at least 2 mm separation between the target and the cord/cauda. More margin is needed if the distance between the isocenter and cord/cauda is larger. It is advisable to place the planning isocenter close to the spinal canal to further mitigate the rotational error.SummaryWe simulated various scenarios of patient setup errors with different rotational combinations of pitch and yaw with 1°, 2°, and 3°, respectively. Rotation was corrected with translation only to mimic the Halcyon treatment scenario. Using the Halcyon for treating a tumor in a single thoracic or lumbar vertebral segment is feasible, but caution should be noted in patients requiring rotational corrections of > 1° in the absence of 6‐DOF correction capabilities.

PRGFlow: Unified SWAP‐aware deep global optical flow for aerial robot navigation

Global optical flow estimation is the foundation stone for obtaining odometry which is used to enable aerial robot navigation. However, such a method has to be of low latency and high robustness whilst also respecting the size, weight, area and power (SWAP) constraints of the robot. A combination of cameras coupled with inertial measurement units (IMUs) has proven to be the best combination in order to obtain such low latency odometry on resource-constrained aerial robots. Recently, deep learning approaches for visual inertial fusion have gained momentum due to their high accuracy and robustness. However, an equally noteworthy benefit for robotics of these techniques are their inherent scalability (adaptation to different sized aerial robots) and unification (same method works on different sized aerial robots). To this end, we present a deep learning approach called PRGFlow for obtaining global optical flow and then loosely fuse it with an IMU for full 6-DoF (Degrees of Freedom) relative pose estimation (which is then integrated to obtain odometry). The network is evaluated on the MSCOCO dataset and the dead-reckoned odometry on multiple real-flight trajectories without any fine-tuning or re-training. A detailed benchmark comparing different network architectures and loss functions to enable scalability is also presented. It is shown that the method outperforms classical feature matching methods by 2 × under noisy data. The supplementary material and code can be found at http://prg.cs.umd.edu/PRGFlow.

Cooperative Transportation of Cable Suspended Payloads With MAVs Using Monocular Vision and Inertial Sensing

Micro Aerial Vehicles (MAVs) have the great potential to be deployed in commercial or health care services such as e-commerce package delivery, transportation of medicines, same-day food delivery, and other time-sensitive transportation tasks. A team of MAVs can cooperatively transport objects to overcome the physical limitations of a single vehicle, while concurrently increasing the system's resilience to vehicles' failures. In this letter, we address the state estimation, control and trajectory tracking problems of cooperative transportation of cable suspended rigid body payloads with MAVs using monocular vision and inertial sensing. The key contributions are (a) a distributed vision-based coordinated control of the cable-suspended rigid body payload on SE(3), (b) a distributed estimation approach that allows each agent to estimate its cable direction and velocity independently, and (c) a new cooperative estimation scheme that can infer the payload's full 6-DoF states. This is obtained by sharing the robots' local position estimates and their relative position with respect to their corresponding attachment points on the payload. It allows to infer the payload's 6-DoF state when it is not directly measurable by each vehicle. The proposed solution runs in real-time on board and is validated through experimental results with multiple quadrotors transporting a rigid body payload via cables.

Can We Localize an Autonomous Vehicle From a Single Image? Deep-Geometric Six Degrees-of-Freedom Localization in Topo-Metric Maps

Abstract We describe a deep-geometric localizer that is able to estimate the full six degrees-of-freedom (DOF) global pose of the camera from a single image in a previously mapped environment. Our map is a topo-metric one, with discrete topological nodes whose 6DOF poses are known. Each topo-node in our map also comprises of a set of points, whose 2D features and 3D locations are stored as part of the mapping process. For the mapping phase, we utilize a stereo camera and a regular stereo visual SLAM pipeline. During the localization phase, we take a single camera image, localize it to a topological node using deep learning, and use a geometric algorithm (perspective-n-point (PnP)) on the matched 2D features (and their 3D positions in the topo map) to determine the full 6DOF globally consistent pose of the camera. Our method divorces the mapping and the localization algorithms and sensors (stereo and mono) and allows accurate 6DOF pose estimation in a previously mapped environment using a single camera. With results in simulated and real environments, our hybrid algorithm is particularly useful for autonomous vehicles (AVs) and shuttles that might repeatedly traverse the same route.

A numerical solver for coupled dynamic simulation of glacial ice impacts considering hydrodynamic-ice-structure interaction

Glacial ice features pose great threats on the safety of ships and offshore structures in the arctic. House sized bergy bits or growlers are of particular concern because of the detection capability limits of marine radars. Analysis and design of structures against collisions from such glacial ice bodies has always been challenging due to the complicated hydrodynamic-ice-structure interaction.This paper proposes a numerical solver for coupled simulation of glacial ice impacts accounting for the effects of hydrodynamic-ice-structure interaction. The solver adopts user subroutines provided in LS-DYNA and combines three different modules, i.e. the BWH (Bressan-Williams-Hill) criterion for the prediction of fracture of steels, a hydrostatic pressure dependent plasticity-based material model for constitutive modelling of ice, and the linear potential flow theory for hydrodynamic loads.The proposed solver is verified and calibrated to ice resistance data from field tests and is then applied to simulate ice collisions on a semi-submersible platform column. Collision scenarios with both in-plane 3DOF and full 6DOF ice motions are considered. The results are discussed with respect to ice motion trajectories, ice crushing and structural damage under the combined action of ice indentation and sliding loads. The dissipated energy predicted by external dynamic models is compared with simulation results and discussed.

Rendezvous in Cis-Lunar Space near Rectilinear Halo Orbit: Dynamics and Control Issues

The paper presents the development of a fully-safe, automatic rendezvous strategy between a passive vehicle and an active one orbiting around the Earth–Moon L2 Lagrangian point. This is one of the critical phases of future missions to permanently return to the Moon, which are of interest to the majority of space organizations. The first step in the study is the derivation of a suitable full 6-DOF relative motion model in the Local Vertical Local Horizontal reference frame, most suitable for the design of the guidance. The main dynamic model is approximated using both the elliptic and circular three-body motion, due to the contribution of Earth and Moon gravity. A rather detailed set of sensors and actuator dynamics was also implemented in order to ensure the reliability of the guidance algorithms. The selection of guidance and control is presented, and evaluated using a sample scenario as described by ESA’s HERACLES program. The safety, in particular the passive safety, concept is introduced and different techniques to guarantee it are discussed that exploit the ideas of stable and unstable manifolds to intrinsically guarantee some properties at each hold-point, in which the rendezvous trajectory is divided. Finally, the rendezvous dynamics are validated using available Ephemeris models in order to verify the validity of the results and their limitations for future more detailed design.

Identification of real translational and rotational displacements of six-axial shakers with only six measured linear accelerations

Qualification tests on the vibrations of on-board equipment depend on the ability of the shaker used to reproduce as precisely as possible the excitation profiles stipulated by standards, given that real operational conditions are generally multiaxial. Using a set of six measured linear accelerations the original method proposed permits determining real six-axial shaker displacements combining three translations and three rotations. Thus, this method, named REDEAT (REal Displacement from Experimental Acceleration with inverse Technique), is useful for introducing these real displacements in the numerical simulation in view of experimentally validating the dynamic behavior model of on-board equipment. Moreover, the REDEAT method evaluates the deviation between real and intended forced displacements. Its principle resides in solving a system of nonlinear time-differential equations expressed, by resorting to numerical integration combining Newmark and Newton–Raphson schemes. In order to overcome the classical divergence issue inherent to the use of numerical integration, data processing tools are employed, consisting of the application of successive high pass filters and window functions. The REDEAT method is applied to a hydraulic 6-DOF shaker equipped with accelerometers and gyroscopes, through a combination of two harmonic rotations and a full 6-DOF random motion. In addition, the influence of the method’s parameters and of the six input accelerograms chosen from twelve is investigated.

Task-oriented optimal dimensional synthesis of robotic manipulators with limited mobility

• The performance of lower-mobility manipulators can be improved with task-oriented designs. • An optimization method is proposed for the dimensional synthesis of these robotic systems . • The proposed algorithm finds the optimal geometry for best accuracy over prescribed paths. • A case study with a 4-DoF robot for high-precision laser operations in aeroengines is reported. In this article, an optimization method is proposed for the dimensional synthesis of robotic manipulators with limited mobility, i.e. with less than 6 degrees-of-freedom (“DoF”), with a prescribed set of tasks in a constrained environment. Since these manipulators cannot achieve full 6-DoF mobility, they are able to follow only certain paths with prescribed position and orientation in space. While the most common approach to this problem employs pure path-planning algorithms, operations in narrow and complex environments might require changes to the robot design too. For this reason, this paper presents an improved approach which aims to minimize position and orientation error with a dimensional synthesis. First, a novel methodology that combines a path planning algorithm and dimensional synthesis has been proposed in order to optimize both robot geometry and pose for a given set of points. Then, the method is validated with a 4-DoF robot for high-precision laser operations in aeroengines as a case study. The example shows that the proposed procedure provides a stable algorithm with a high convergence rate and a short time to solution for robots with limited mobility in highly constrained scenarios.

CFD Simulations and Experiments of a Submarine in Turn, Zigzag, and Surfacing Maneuvers

We present simulations and experiments of the generic submarine Joubert BB2 performing standard turn, zigzag, and surfacing maneuvers in calm water at depth. The free-sailing experiments, performed at Maritime Research Institute Netherlands (MARIN), are unique in that they present an open dataset for the community to benchmark maneuvering prediction methodologies. Computations were performed with explicitly gridded sailplanes, tail planes, and propellers using a dynamic overset technique. This study analyzes a 20-degree turning maneuver with vertical control commanding the stern planes and a 20/20 zigzag maneuver with vertical control commanding both sail and stern planes, both of them at a nominal speed of 10 knots, and a 20-degree rise maneuver with horizontal control at 12 knots. The results show that computational fluid dynamics can predict well motions and speeds for free-sailing conditions, but controller commands are harder to replicate. Computations of the rise maneuver with surfacing compare well with experiments, including a crashback maneuver to stop the submarine. Introduction Simulation of free-running submarine maneuvers is demanding in terms of computational fluid dynamics (CFD) capabilities and computational resources. Full 6-DoF motions, moving appendages, control algorithms, and scalable performance are some of the requirements to execute free-running maneuvering simulations. Open literature data for validation of submarine maneuvers are very limited and consist mostly of static conditions (see for instance rotating arm [Toxopeus et al. 2012] and static drift [Roddy 1990; Toxopeus 2008] data for Defense Advanced Research Projects Agency (DARPA) Suboff and static drift [Quick &amp; Woodyatt 2014] for the Joubert hull form [Joubert 2006]). A new dataset has been recently made available to the research community for the generic submarine model BB2 (Overpelt et al. 2015), based on the Joubert hull form.

HDMI-Loc: Exploiting High Definition Map Image for Precise Localization via Bitwise Particle Filter

In this letter, we propose a method for accurately estimating the 6-Degree Of Freedom (DOF) pose in an urban environment when a High Definition (HD) map is available. An HD map expresses 3D geometric data with semantic information in a compressed format and thus is more memory-efficient than point cloud maps. The small capacity of HD maps can be a significant advantage for autonomous vehicles in terms of map storage and updates within a large urban area. Unfortunately, existing approaches failed to sufficiently exploit HD maps by only estimating partial pose. In this study, we present a full 6-DOF localization against an HD map using an onboard stereo camera with semantic information from roads. We introduce an 8-bit representation for road information, which allow for effective bitwise operation when matching between query data and the HD map. For the pose estimation, we leverage a particle filter followed by a full 6-DOF pose optimization. Our experimental results show a median error of approximately 0.3 m in the lateral and longitudinal directions for a drive of approximately 11 km. These results can be used by autonomous vehicles to correct the global position without using Global Positioning System (GPS) data in highly complex urban environments. The median operation speed is approximately 60 msec supporting 10 Hz.

Vision-Based Robotic Traversal of Textureless Smooth Surfaces

This investigation focuses on servoing a robot's tool with respect to a smooth workpiece surface by making use of the surface's characteristic local differential properties. A novel formulation for 6 degree-of-freedom (DoF) textureless visual servoing based on these properties is proposed, which extends an existing 3-DoF scheme. Our approach naturally combines the geometric tools of computer-aided design and machining (CAD/CAM) theory with the manipulator control tools of visual servoing synergistically to achieve full 6-DoF pose control. A novel family of observed feature sets and their associated interaction matrices are presented. A geometric condition on the surface shape is derived under which local asymptotic stability for 6-DoF is guaranteed. Validation of the proposed method is performed in simulation and experiment using an articulated desktop robot equipped with only a monocular camera and 16 laser pointers.

Open-Loop Orientation Control Using Dynamic Magnetic Fields

Remote magnetic control of soft magnetic objects has been limited to 2D orientation and 3D position. In this letter, we extend the five degree-of-freedom (5-DoF) control approach to full 6-DoF. We prove that 6-DoF control is possible for objects that have an apparent magnetic susceptibility tensor with unique eigenvalues. We further show that the object’s orientation can be specified with a dynamic magnetic field and can be controlled without orientation feedback.The theory is demonstrated by rotating a soft magnetic object about each of its principle axes using a metronome like dynamic field.

Real-Time Multi-person Motion Capture from Multi-view Video and IMUs

A real-time motion capture system is presented which uses input from multiple standard video cameras and inertial measurement units (IMUs). The system is able to track multiple people simultaneously and requires no optical markers, specialized infra-red cameras or foreground/background segmentation, making it applicable to general indoor and outdoor scenarios with dynamic backgrounds and lighting. To overcome limitations of prior video or IMU-only approaches, we propose to use flexible combinations of multiple-view, calibrated video and IMU input along with a pose prior in an online optimization-based framework, which allows the full 6-DoF motion to be recovered including axial rotation of limbs and drift-free global position. A method for sorting and assigning raw input 2D keypoint detections into corresponding subjects is presented which facilitates multi-person tracking and rejection of any bystanders in the scene. The approach is evaluated on data from several indoor and outdoor capture environments with one or more subjects and the trade-off between input sparsity and tracking performance is discussed. State-of-the-art pose estimation performance is obtained on the Total Capture (mutli-view video and IMU) and Human 3.6M (multi-view video) datasets. Finally, a live demonstrator for the approach is presented showing real-time capture, solving and character animation using a light-weight, commodity hardware setup.

SCOUT3D – AN UNDERWATER LASER SCANNING SYSTEM FOR MOBILE MAPPING

Abstract. This paper presents an underwater laser scanning system and GNSS based trajectory estimation system for scanning from a surface vessle in shallow water. The system has an above-the-water and an underwater component. Above-the-water two low-cost multiband GNSS receivers with an antenna baseline of one meter are used for RTK positioning with heading. The full 6-DOF is estimated by fusing the satellite navigation data with a MEMS-based INS. The 3D data is captured in water using a structured light scanner consisting of a low-light underwater camera and a green cross line laser projector. We describe the development of the system and employed hardware components. We show results of scanning a large test object in a water tank acquired by from a tripod with a motorized yaw axis. Additionally, we demonstrate first results of mobile mapping from a floating platform. We evaluate the performance of the system by measuring the 6-DOF trajectory with an external optical tracking system. Additionally, we assess the quality of the created point cloud using reference objects placed in the scene.