On-Board Dual-Stereo-Vision for the Navigation of an Autonomous MAV

Schauwecker, Konstantin; Zell, Andreas

doi:10.1007/s10846-013-9907-6

Cited by 51 publications

(36 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Stereo vision can provide depth information and has been used in a 4 g flapping-wing vehicle to reactively avoid obstacles 87 and in larger quadcopters to perform simultaneous localization and mapping 88 . Furthermore, both simultaneous localization and mapping and altitude control have been demonstrated in quadcopters with two sets of stereo cameras, one set pointing forwards and one set pointing downwards 89 . Simultaneous localization and mapping algorithms have also been combined with optic flow methods to estimate distances from the surrounding environment and stabilize the drone 90 .…”

Section: Cognitive Autonomymentioning

confidence: 99%

Science, technology and the future of small autonomous drones

Floreano

Wood

2015

Nature

987

503

View full text Add to dashboard Cite

Drones, a popular nickname for unmanned aerial vehicles and micro aerial vehicles, often conjure up images of unmanned aeroplanes that fly thousands of miles for espionage and to deploy munitions. However, over the past few years, an increasing number of public and private research laboratories have been working on small, human-friendly drones that one day may autonomously fly in confined spaces and in close proximity to people. The development of these small drones, which is the main focus of this Review, has been supported by the miniaturization and cost reduction of electronic components (microprocessors, sensors, batteries and wireless communication units), largely driven by the portable electronic device industry. These improvements have enabled the prototyping and commercialization of small (typically less than 1 kg) drones at smartphone prices.Small drones will have important socio-economic impacts (Fig. 1). Images from drones that are capable of flying a few metres above the ground will fill a gap between expensive, weather-dependent and lowresolution images provided by satellites and car-based images limited to human-level perspectives and the availability of accessible roads. Specialized flying cameras and cloud-based data analytics will allow farmers to continuously monitor the quality of crop growth. Such platforms will enable construction companies to measure work progress in real time. Drones will let mining companies obtain precise volumetric data of excavations. Energy and infrastructure companies will be able to exhaustively survey pipelines, roads and cables. Humanitarian organizations could immediately assess and adapt aid efforts in continuously changing refugee camps. Transportation drones that are capable of safely taking off and landing in the proximity of buildings and humans will allow developing countries -without a suitable road network -to rapidly deliver goods and to finally unleash the full potential of their e-commerce telecommunication infrastructure. Transportation drones will also help developed countries to improve the quality of service in congested or remote areas, and will enable rescue organizations to quickly deliver medical supplies in the field and on demand. Inspection drones that are capable of flying in confined spaces will help fire-fighting and emergency units to assess dangers faster and more safely, logistic companies to detect cracks in the inner and outer shells of ships, road maintenance companies to measure signs of wear and tear in bridges and tunnels, security companies to improve building safety by monitoring areas outside the range of surveillance cameras, and disaster mitigation agencies to inspect partially collapsed buildings where ground clutter is an obstacle for terrestrial robots. Coordinated teams of autonomous drones will enable missions that last longer than the flight time of a single drone by allowing some drones to temporarily leave the team for battery replacement. Drone teams will permit rescue organizations to quickly deploy dedicated commu...

show abstract

Section: Cognitive Autonomymentioning

confidence: 99%

Science, technology and the future of small autonomous drones

Floreano

Wood

2015

Nature

987

503

View full text Add to dashboard Cite

show abstract

“…We then look at vision-based MAVs that run real-time pose estimation algorithms on-board. Schauwecker and Zell [27] deploy one downward-looking stereo camera and one forwardlooking stereo camera on a MAV, independently estimate the MAV's pose from each camera, and fuse both pose estimates. In contrast, we use all cameras to obtain a single pose estimate.…”

Section: Related Workmentioning

confidence: 99%

Self-calibration and visual SLAM with a multi-camera system on a micro aerial vehicle

2015

View full text Add to dashboard Cite

Abstract-The use of a multi-camera system enables a robot to obtain a surround view, and thus, maximize its perceptual awareness of its environment. An accurate calibration is a necessary prerequisite if vision-based simultaneous localization and mapping (vSLAM) is expected to provide reliable pose estimates for a micro aerial vehicle (MAV) with a multi-camera system. On our MAV, we set up each camera pair in a stereo configuration. We propose a novel vSLAM-based self-calibration method for a multi-sensor system that includes multiple calibrated stereo cameras and an inertial measurement unit (IMU). Our selfcalibration estimates the transform with metric scale between each camera and the IMU. Once the MAV is calibrated, the MAV is able to estimate its global pose via a multi-camera vSLAM implementation based on the generalized camera model. We propose a novel minimal and linear 3-point algorithm that uses inertial information to recover the relative motion of the MAV with metric scale. Our constant-time vSLAM implementation with loop closures runs on-board the MAV in real-time. To the best of our knowledge, no published work has demonstrated realtime on-board vSLAM with loop closures. We show experimental results in both indoor and outdoor environments. The code for both the self-calibration and vSLAM is available as a set of ROS packages at https://github.com/hengli/vmav-ros-pkg. I. INTRODUCTIONVision-based MAVs are more versatile than laser-based MAVs. Whereas a laser only provides geometry data, a camera can provide both geometry data via stereo and structure-frommotion techniques, and appearance data. A camera is a passive sensor while a laser is an active sensor and is thus susceptible to interference. Furthermore, a camera is lighter and has a smaller footprint. However, utmost care has to be taken when choosing the camera configuration for a vision-based MAV expected to operate robustly in challenging environments. A single-camera configuration introduces limited perceptual awareness, and in turn, flight constraints because if the camera observes too few features for some time, localization can fail and lead to a crash. In the case of a single downward-looking camera [32], the MAV cannot fly too close to the ground which often has little texture, and at the same time, it cannot perform obstacle avoidance due to the absence of a forward-looking camera. In the case of a forward-looking camera, constraints are imposed on the path planning. For example, in [24], any planned path ensures that there are a sufficient number of features for localization. In addition, in [11], a path is planned such that the MAV does not move outside the camera's field of view, and inadvertently crash into an unseen obstacle.In this paper, we use a multi-camera system on a MAV; together with the use of fish-eye lenses, this camera configuration provides a surround view of the vicinity. Our multi-camera

show abstract

“…Obstacle avoidance is often neglected, e.g., by flying in a sufficient height when autonomously flying between waypoints. Most approaches to obstacle avoidance for MAVs are camera-based, due to the limited payload [10], [11], [12], [13], [14], [15], [16], [17]. Hence, collision avoidance is restricted to the field of view (FoV) of the cameras.…”

Section: Related Workmentioning

confidence: 99%

Obstacle detection and navigation planning for autonomous micro aerial vehicles

Nieuwenhuisen

Droeschel

Beul

et al. 2014

2014 International Conference on Unmanned Aircraft Systems (ICUAS)

View full text Add to dashboard Cite

Abstract-Obstacle detection and real-time planning of collision-free trajectories are key for the fully autonomous operation of micro aerial vehicles in restricted environments.In this paper, we propose a complete system with a multimodal sensor setup for omnidirectional obstacle perception consisting of a 3D laser scanner, two stereo camera pairs, and ultrasonic distance sensors. Detected obstacles are aggregated in egocentric local multiresolution grid maps. We generate trajectories in a multi-layered approach: from mission planning to global and local trajectory planning, to reactive obstacle avoidance.We evaluate our approach in simulation and with the real autonomous micro aerial vehicle.

show abstract

On-Board Dual-Stereo-Vision for the Navigation of an Autonomous MAV

Cited by 51 publications

References 23 publications

Science, technology and the future of small autonomous drones

Science, technology and the future of small autonomous drones

Self-calibration and visual SLAM with a multi-camera system on a micro aerial vehicle

Obstacle detection and navigation planning for autonomous micro aerial vehicles

Contact Info

Product

Resources

About