3D Automatic Target Recognition for UAV Platforms

Kechagias‐Stamatis, Odysseas; Aouf, Nabil; Nam, David

doi:10.1109/sspd.2017.8233223

Cited by 10 publications

(10 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The suggested architecture performs well for the following reasons. First, HoD‐S is robust to highly sparse point clouds [21, 22, 25] providing to the adaptive H ∞ filter only well‐established correspondences. Second, the H ∞ filter has been designed for robustness against extreme nuisances, and third, the adaptive measurement noise covariance H k affords further performance improvement over the standard H ∞ recursive filter.…”

Section: Methodsmentioning

confidence: 99%

“…We describe all vertices belonging to each point cloud P k and P k +1 using a variant of the histogram of distances (HoD) [20] entitled HoD‐Short (HoD‐S) [21, 22]. Despite current literature offering quite a few 3D local feature descriptors such as the fast point feature histogram [23], rotational projection statistics [17] and signatures of histograms of orientations [24], we used the HoD‐S due to its processing efficiency and robustness to highly sparse point clouds [21, 22, 25] as examined in this work.…”

Section: H∞ Lidar Odometrymentioning

confidence: 99%

See 1 more Smart Citation

H ∞ LIDAR odometry for spacecraft relative navigation

Kechagias‐Stamatis

Aouf

2019

IET Radar, Sonar & Navigation

Self Cite

View full text Add to dashboard Cite

Current Light Detection and Ranging (LIDAR) based odometry solutions that are used for spacecraft relative navigation suffer from quite a few deficiencies. These include an off-line training requirement and relying on the Iterative Closest point (ICP) that does not guarantee a globally optimum solution. To encounter this, we suggest a robust architecture that overcomes the problems of current proposals by combining the concepts of 3D local feature matching with an adaptive variant of the H∞ recursive filtering process. Trials on real laser scans of an EnviSat model demonstrate that the proposed architecture affords at least one order of magnitude better accuracy compared to ICP.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: H∞ Lidar Odometrymentioning

confidence: 99%

H ∞ LIDAR odometry for spacecraft relative navigation

Kechagias‐Stamatis

Aouf

2019

IET Radar, Sonar & Navigation

Self Cite

View full text Add to dashboard Cite

show abstract

“…The parameters of the architecture and of the competing methods are tuned based on the SLA Scenario. Table 2 presents the tuned parameters, while the ones not tuned are fixed either to the ones originally proposed by their authors or for OUR-CVFH and Spin Images to their PCL implementation [34,49,50]. Odometry performance is evaluated based on drift, i.e.…”

Section: Methodsmentioning

confidence: 99%

“…Alternatives to pure ICP registration for pose estimation have also been proposed by substituting ICP with a UKF filter, an iterative least-squares (LS) scheme, or with an Extended Kalman Filter (EKF) [32], [33]. An additional alternative is suggested in [18] that combines 3D local feature matching based on the Histogram of Distances -Short (HoD-S) descriptor [34] and the H∞ filter.…”

Section: Introductionmentioning

confidence: 99%

High-speed multi-dimensional relative navigation for uncooperative space objects

2019

View full text Add to dashboard Cite

This work proposes a high-speed Light Detection and Ranging (LIDAR) based navigation architecture that is appropriate for uncooperative relative space navigation applications. In contrast to current solutions that exploit 3D LIDAR data, our architecture transforms the odometry problem from the 3D space into multiple 2.5D ones and completes the odometry problem by utilizing a recursive filtering scheme. Trials evaluate several current state-of-theart 2D keypoint detection and local feature description methods as well as recursive filtering techniques on a number of simulated but credible scenarios that involve a satellite model developed by Thales Alenia Space (France). Most appealing performance is attained by the 2D keypoint detector Good Features to Track (GFFT) combined with the feature descriptor KAZE, that are further combined with either the H∞ or the Kalman recursive filter. Experimental results demonstrate that compared to current algorithms, the GFTT/ KAZE combination is highly appealing affording one order of magnitude more accurate odometry and a very low processing burden, which depending on the competitor method, may exceed one order of magnitude faster computation.

show abstract

“…Investigations involve solutions based on numerous spatial, i.e. 2D/ 3D and data domains, such as 2D infrared (IR) [1][2][3][4][5] , 2D Synthetic Aperture Radar (SAR) [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21] , 2D Inverse SAR (ISAR) 22 and 3D Light Detection and Ranging (LIDAR) [23][24][25][26][27] , with each of these data modalities having its own strengths and weaknesses. For example, state-of-the-art local feature (data) descriptors from the visual domain have already proven their capabilities in the IR domain, but IR suffers from the time of day and the target's history 28 .…”

Section: Introductionmentioning

confidence: 99%

Target recognition for synthetic aperture radar imagery based on convolutional neural network feature fusion

Kechagias‐Stamatis

2018

J. Appl. Rem. Sens.

Self Cite

View full text Add to dashboard Cite

Driven by the great success of deep Convolutional Neural Networks (CNNs) that are currently used by quite a few computer vision applications, we extend the usability of visual based CNNs into the Synthetic Aperture Radar (SAR) data domain without employing transfer learning. Our SAR Automatic Target Recognition (ATR) architecture efficiently extends the pre-trained Visual Geometry Group CNN from the visual domain into the X-band SAR data domain by clustering its neuron layers, bridging the visual-SAR modality gap by fusing the features extracted from the hidden layers and by employing a local feature matching scheme. Trials on the moving and stationary target acquisition dataset under various setups and nuisances demonstrate a highly appealing ATR performance gaining 100% and 99.79% in the 3-class and 10-class ATR problem respectively. We also confirm the validity, robustness and conceptual coherence of the proposed method by extending it to several state-of-the-art CNNs and commonly used local feature similarity/ match metrics.

show abstract

3D Automatic Target Recognition for UAV Platforms

Cited by 10 publications

References 19 publications

H ∞ LIDAR odometry for spacecraft relative navigation

H ∞ LIDAR odometry for spacecraft relative navigation

High-speed multi-dimensional relative navigation for uncooperative space objects

Target recognition for synthetic aperture radar imagery based on convolutional neural network feature fusion

Contact Info

Product

Resources

About