Real time monocular visual odometry using Optical Flow: Study on navigation of quadrotors UAV

Mansur, Syaiful; Ḥabīb, Muḥammad Ibn; Pratama, Gilang Nugraha Putu; Cahyadi, Adha Imam; Ardiyanto, Igi

doi:10.1109/icstc.2017.8011864

Cited by 9 publications

(3 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional computer vision algorithms propose the use of different feature extractors such as ORB descriptors [16] in which a 3D dense map is generated with the features extracted. Also, the optical flow has been combined with extractors to improve estimation [2,12,13].…”

Section: Related Workmentioning

confidence: 99%

LatinX in AI at Computer Vision and Pattern Recognition Conference 2022

Cocoma-Ortega¹,

Martínez-Carranza²

2022

View full text Add to dashboard Cite

Autonomous navigation is a challenging task that requires solving individual problems such as the camera pose. Even more, if agile motion is performed in the navigation, the problem becomes complex due to the necessity to know the pose (trajectory generated) as fast as possible to continue with the agile motion. Several works have proposed approaches based on geometric algorithms and deep learning-based solutions that reduce error in estimation. Still, the prediction is performed at 30Hz on average in most cases. It is desirable to increase the frequency of operation to allow cameras with higher frame rates to capture the motion correctly. Motivated by the latter, we propose a two-stream network that allows predicting pose at a frame rate up to 78f ps with an average relative error of 2.21m.

show abstract

Section: Related Workmentioning

confidence: 99%

LatinX in AI at Computer Vision and Pattern Recognition Conference 2022

Cocoma-Ortega¹,

Martínez-Carranza²

2022

View full text Add to dashboard Cite

show abstract

“…For instance, Flowdometry [41] estimates displacements and rotations from OF-input. Mansur et al [37] proposed a VO-based dead reckoning system that uses OF to match features. Zhao et al [59] combined two CNNs to estimate the VO-motion: FlowNet2-ss [18] estimates the OF and PCNN [11] links two images to process global and local pose information.…”

Section: Optical Flowmentioning

confidence: 99%

ViPR: Visual-Odometry-aided Pose Regression for 6DoF Camera Localization

Ott,

Feigl,

Löffler

et al. 2019

Preprint

View full text Add to dashboard Cite

Visual Odometry (VO) accumulates a positional drift in long-term robot navigation tasks. Although Convolutional Neural Networks (CNNs) improve VO in various aspects, VO still suffers from moving obstacles, discontinuous observation of features, and poor textures or visual information. While recent approaches estimate a 6DoF pose either directly from (a series of) images or by merging depth maps with the optical flow (OF), research that combines absolute pose regression with OF is limited.We propose ViPR, a novel architecture for long-term 6DoF VO that leverages synergies between absolute pose estimates (from PoseNet-like architectures) and relative pose estimates (from FlowNet-based architectures) by combining both through recurrent layers. Experiments with known publicly available datasets and with our own Industry dataset show that our novel design outperforms existing techniques in long-term navigation tasks.

show abstract

“…Depending on the number of available cameras, VO can be monocular or stereo. Furthermore, it can be based on different image processing techniques: optical flow (Mansur et al, 2017), features matching (Chuanqi et al, 2017), or feature tracking (Johnson et al, 2008). (Mostafa et al, n.d.) proposed a novel approach based on integrating a monocular camera, IMU, and Artificial Intelligent (AI) through EKF, where the AI is used to estimate the underlying function of the camera and diminish the need for estimating the geometric camera model.…”

Section: Introductionmentioning

confidence: 99%

Uavs Enhanced Navigation in Outdoor GNSS Denied Environment Using Uwb and Monocular Camera Systems

Zahran

Masiero

Mostafa

et al. 2019

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

<p><strong>Abstract.</strong> The demand for small Unmanned Aerial Vehicles (UAVs) is massively increasing these days, due to the wide variety of applications utilizing such vehicles to perform tasks that may be dangerous or just to save time, effort, or cost. Small UAVs navigation system mainly depends on the integration between Global Navigation Satellite Systems (GNSS) and Inertial Measurement Unit (INS) to estimate the Positions, Velocities, and Attitudes (PVT) of the vehicle. Without GNSS such UAVs cannot navigate for long periods of time depending on INS alone, as the low-cost INS typically exhibits massive accumulation of errors during GNSS absence. Given the importance of ensuring full operability of the UAVs even during GNSS signals unavailability, other sensors must be used to bound the INS errors and enhance the navigation system performance. This paper proposes an enhanced UAV navigation system based on integration between monocular camera, Ultra-Wideband (UWB) system, and INS. In addition to using variable EKF weighting scheme. The paper also investigates this integration in the case of low density of UWB anchors, to reduce the cost required for such UWB system infrastructure. A GoPro Camera and UWB rover were attached to the belly of a quadcopter, an on the shelf commercial drone (3DR Solo), during the experimental flight. The velocity of the vehicle is estimated with Optical Flow (OF) from camera successive images, while the range measurements between the UWB rover and the stationary UWB anchors, which were distributed on the field, were used to estimate UAV position.</p>

show abstract

Real time monocular visual odometry using Optical Flow: Study on navigation of quadrotors UAV

Cited by 9 publications

References 18 publications

LatinX in AI at Computer Vision and Pattern Recognition Conference 2022

LatinX in AI at Computer Vision and Pattern Recognition Conference 2022

ViPR: Visual-Odometry-aided Pose Regression for 6DoF Camera Localization

Uavs Enhanced Navigation in Outdoor GNSS Denied Environment Using Uwb and Monocular Camera Systems

Contact Info

Product

Resources

About