Analysis of Simulated Imagery for Real-Time Vision-Based Automated Aerial Refueling

Parsons, Christopher A.; Paulson, Zachary; Nykl, Scott; Dallman, William; Woolley, Brian G.; Pecarina, John M.

doi:10.2514/1.i010658

Cited by 14 publications

(12 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This section explains the experiments that were conducted to validate our previously proposed computer vision pipeline. The base pipeline comes from Parsons et al's previous work [1] and functions as follows:…”

Section: Ground Experiments Design and Resultsmentioning

confidence: 99%

“…In this work, we use geometrically accurate models, high-quality textures, and realistic lighting to replicate real refueling approaches and generate synthetic imagery. This is the same simulation environment that several researchers have used [1,32,33] for their AAR experiments. The cameras in the simulation have the same resolution and field of view as their physical counterparts used in the ground experiment.…”

Section: A Computer Simulationmentioning

confidence: 99%

“…Due to aerial refueling mission constraints, we cannot modify existing aircraft or use Global Positioning System (GPS) signals to augment a vision-based approach. Parsons et al have previously proposed a method for solving this problem using a stereo vision system [1]. Here, we seek to validate and improve that system.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Long-Range Pose Estimation for Aerial Refueling Approaches Using Deep Neural Networks

Lee

Dallmann

Nykl

et al. 2020

Journal of Aerospace Information Systems

Self Cite

View full text Add to dashboard Cite

Automated aerial refueling (AAR) provides unique challenges for computer vision systems. Aerial refueling maneuvers require high-precision low-variance pose estimates. The performance of two stereoscopic (stereo) vision systems is quantified in ground tests specially designed to mimic AAR. In this experiment, three-dimensional (3-D) pose-estimation errors of 6 cm on a target 30 m from the current vision system are achieved. Next, a novel computer vision pipeline to efficiently generate a 3-D point cloud of the target object using stereo vision that leverages a convolutional neural network (CNN) is proposed. Using the proposed approach, a high-fidelity 3-D point cloud with ultra-high-resolution imagery 11.3 times faster than previous approaches can be generated.

show abstract

Section: Ground Experiments Design and Resultsmentioning

confidence: 99%

Section: A Computer Simulationmentioning

confidence: 99%

See 1 more Smart Citation

Long-Range Pose Estimation for Aerial Refueling Approaches Using Deep Neural Networks

Lee

Dallmann

Nykl

et al. 2020

Journal of Aerospace Information Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…The Kalman filter is used to accurately estimate the pose of the drogue [28]. By analyzing the motion law between the drogue and the UAV [15], knowing that the UAV approaches the drogue at a uniform speed in the z-axis direction. The drogue swings in a certain frequency range on the X and Y axes direction due to the presence of atmospheric turbulence.…”

Section: ) Principle Of Spatial Circular Pose Solvingmentioning

confidence: 99%

“…Johnson et al [14] designed an extended Kalman filter to combine the stereo camera system and inertial navigation for automated aerial refueling. Parsons et al [15] presents a 3D graphical simulation that replicates a complete aerial refueling scenario. Luo et al [16] stated a binocular vision system that is based on the light-emitting diodes (LEDs) features.…”

Section: Introductionmentioning

confidence: 99%

A Novel Method for Measuring Drogue-UAV Relative Pose in Autonomous Aerial Refueling Based on Monocular Vision

Zhao

Liu

et al. 2019

IEEE Access

View full text Add to dashboard Cite

The key to the docking phase of autonomous aerial refueling missions is the relative pose measurement between the drogue and unmanned aerial vehicles. A novel measurement method for drogue-UAV (unmanned aerial vehicle) relative poses based on monocular vision is proposed in this paper. An adaptive arc-level structural feature extraction algorithm is applied to obtain the features that can be used for the measurements. In this algorithm, the projection of the end plane of the drogue in the image plane is accurately extracted based on an extraction quality metric. The projection contour and centroid of the internal black part of the drogue form the structural feature for the measurements. A robust pose estimation algorithm is presented to estimate the relative pose between the drogue and the UAV. The pose can be solved using a cone that is composed of the optical center of the camera and the structural feature. A Kalman filter is then applied to robustly estimate the pose of the drogue. A simulation and ground experiment are used to verify the robustness and effectiveness of the proposed method, which achieves good performance compared with other methods.

show abstract

Augmenting Flight Imagery from Aerial Refueling

Anderson

Nykl

Wischgoll

2019

Advances in Visual Computing

View full text Add to dashboard Cite

Analysis of Simulated Imagery for Real-Time Vision-Based Automated Aerial Refueling

Cited by 14 publications

References 19 publications

Long-Range Pose Estimation for Aerial Refueling Approaches Using Deep Neural Networks

Long-Range Pose Estimation for Aerial Refueling Approaches Using Deep Neural Networks

A Novel Method for Measuring Drogue-UAV Relative Pose in Autonomous Aerial Refueling Based on Monocular Vision

Augmenting Flight Imagery from Aerial Refueling

Contact Info

Product

Resources

About