Vision-based autonomous hovering for a miniature quad-rotor

Gomez-Balderas, Jose-Ernesto; Salazar, Sergio; Guerrero, José Alfredo; Lozano, Rogelio

doi:10.1017/s0263574713000611

Cited by 7 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Loianno et al have shown that the very promising embedded software developed in the framework of the tango project could lead to an efficient stabilization of a quadrotor as well as creating 3D maps ( [10]). Recent studies have proposed various vision-based control strategies to make a quadrotor able to follow a line ( [11], [12]), to hover ( [13], [14]), land automatically in a safe place ( [15]) and track a moving robotic platform with active markers ( [16], [17]). Unlike studies based on a fixed camera, we propose here a new approach where a novel visual sensor allows a small quadrotor to locate and track a non-emissive moving target with only 24 pixels.…”

Section: Introductionmentioning

confidence: 99%

A novel hyperacute gimbal eye to implement precise hovering and target tracking on a quadrotor

Manecy

Diperi

Boyron

et al. 2016

2016 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

This paper presents a new minimalist bio-inspired artificial eye of only 24 pixels, able to locate accurately a target placed in its small field of view (±10°). The eye is mounted on a very light custom-made gimbal system which makes the eye able to track faithfully a moving target. We have shown, that our gimbal eye can be embedded on a small quadrotor to achieve accurate hovering with respect to a target placed onto the ground. Our aiborne eye was enhanced with a bioinspired reflex in charge of locking efficiently the robot's gaze onto a target and compensate for the robot's rotations and disturbances. The use of very few pixels allowed to implement a visual processing algorithm at a refresh rate of 400 Hz. This high refresh rate coupled to a very fast control of the eye's orientation allowed the robot to track a target moving at a speed up to 200°• s −1 .

show abstract

Section: Introductionmentioning

confidence: 99%

A novel hyperacute gimbal eye to implement precise hovering and target tracking on a quadrotor

Manecy

Diperi

Boyron

et al. 2016

2016 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

show abstract

“…This is the main strategy used in sensor-fusion systems. Similarly to Gomez-Balderas et al [7], Saripalli and Sukhatme [8] combined vision with differential GPS for estimating the relative pose of the UAV with respect to the helipad. In this work, geometric invariant extraction is used to identify a H-shape landing pad.…”

Section: Related Workmentioning

confidence: 99%

Autonomous Vehicular Landings on the Deck of an Unmanned Surface Vehicle using Deep Reinforcement Learning

Polvara

Sharma²,

Wan³

et al. 2019

Robotica

View full text Add to dashboard Cite

SummaryAutonomous landing on the deck of a boat or an unmanned surface vehicle (USV) is the minimum requirement for increasing the autonomy of water monitoring missions. This paper introduces an end-to-end control technique based on deep reinforcement learning for landing an unmanned aerial vehicle on a visual marker located on the deck of a USV. The solution proposed consists of a hierarchy of Deep Q-Networks (DQNs) used as high-level navigation policies that address the two phases of the flight: the marker detection and the descending manoeuvre. Few technical improvements have been proposed to stabilize the learning process, such as the combination of vanilla and double DQNs, and a partitioned buffer replay. Simulated studies proved the robustness of the proposed algorithm against different perturbations acting on the marine vessel. The performances obtained are comparable with a state-of-the-art method based on template matching.

show abstract

“…Vision combined with structured environments has been applied successfully in hovering control [ 4 , 5 , 6 ]. For instance, an autonomous quadrotor aircraft can perform stable hovering above a pattern glyph with a data-fusion algorithm using both visual system measurements and inertial sensors [ 6 ]. However, the vision system relied on the pattern glyph, which is specially designed to align the aircraft with the glyph.…”

Section: Introductionmentioning

confidence: 99%

Monocular-Vision-Based Autonomous Hovering for a Miniature Flying Ball

Lin

Han

Qing-sheng

2015

Sensors

View full text Add to dashboard Cite

This paper presents a method for detecting and controlling the autonomous hovering of a miniature flying ball (MFB) based on monocular vision. A camera is employed to estimate the three-dimensional position of the vehicle relative to the ground without auxiliary sensors, such as inertial measurement units (IMUs). An image of the ground captured by the camera mounted directly under the miniature flying ball is set as a reference. The position variations between the subsequent frames and the reference image are calculated by comparing their correspondence points. The Kalman filter is used to predict the position of the miniature flying ball to handle situations, such as a lost or wrong frame. Finally, a PID controller is designed, and the performance of the entire system is tested experimentally. The results show that the proposed method can keep the aircraft in a stable hover.

show abstract

Vision-based autonomous hovering for a miniature quad-rotor

Cited by 7 publications

References 28 publications

A novel hyperacute gimbal eye to implement precise hovering and target tracking on a quadrotor

A novel hyperacute gimbal eye to implement precise hovering and target tracking on a quadrotor

Autonomous Vehicular Landings on the Deck of an Unmanned Surface Vehicle using Deep Reinforcement Learning

Monocular-Vision-Based Autonomous Hovering for a Miniature Flying Ball

Contact Info

Product

Resources

About