Efficient Object Detection Model for Real-Time UAV Applications.
作者： Ali Jannesari , Chandan Kumar , Subrahmanyam Vaddi
DOI:
关键词:
摘要: Unmanned Aerial Vehicles (UAVs) especially drones, equipped with vision techniques have become very popular in recent years, their extensive use wide range of applications. Many these applications require computer techniques, particularly object detection from the information captured by on-board camera. In this paper, we propose an end to model running on a UAV platform which is suitable for real-time We deep feature pyramid architecture makes inherent properties features extracted Convolutional Networks capturing more generic images (such as edge, color etc.) along minute detailed specific classes contained our problem. VisDrone-18 dataset studies contain different objects such pedestrians, vehicles, bicycles etc. provide software and hardware used study. implemented both ResNet MobileNet convolutional bases. Our combined modified focal loss function, produced desirable performance 30.6 mAP inference time 14 fps. compared results RetinaNet-ResNet-50 HAL-RetinaNet shown that backend extractor gave best terms accuracy, speed memory efficiency real drones.
arxiv.org参考文章(22)
Shaoqing Ren, Kaiming He, Ross Girshick, Jian Sun, Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 39, pp. 1137- 1149 ,(2017) , 10.1109/TPAMI.2016.2577031
Patrick Doherty, Piotr Rudol, A UAV search and rescue scenario with human body detection and geolocalization australasian joint conference on artificial intelligence. pp. 1- 13 ,(2007) , 10.1007/978-3-540-76928-6_1
Hyon Lim, Sudipta N. Sinha, Monocular Localization of a moving person onboard a Quadrotor MAV international conference on robotics and automation. pp. 2182- 2189 ,(2015) , 10.1109/ICRA.2015.7139487
Christian Forster, Matthias Faessler, Flavio Fontana, Manuel Werlberger, Davide Scaramuzza, Continuous on-board monocular-vision-based elevation mapping applied to autonomous landing of micro aerial vehicles international conference on robotics and automation. pp. 111- 118 ,(2015) , 10.1109/ICRA.2015.7138988
Lorenz Meier, Petri Tanskanen, Lionel Heng, Gim Hee Lee, Friedrich Fraundorfer, Marc Pollefeys, PIXHAWK: A micro aerial vehicle design for autonomous flight using onboard computer vision Autonomous Robots. ,vol. 33, pp. 21- 39 ,(2012) , 10.1007/S10514-012-9281-4
Alessandro Renzaglia, Stephan Weiss, Petri Tanskanen, Roland Siegwart, Jan Carsten Stumpf, Simon Lynen, Markus Wilhelm Achtelik, Laurent Kneip, Davide Scaramuzza, Gim Hee Lee, Lefteris Doitsidis, Friedrich Fraundorfer, Savvas A Chatzichristofis, Margarita Chli, Elias Kosmatopoulos, Chiara Troiani, Lorenz Meier, Marc Pollefeys, Michael Achtelik, Lionel Heng, Agostino Martinelli, Daniel Gurdan, None, Vision-Controlled Micro Flying Robots: From System Design to Autonomous Navigation and Mapping in GPS-Denied Environments IEEE Robotics & Automation Magazine. ,vol. 21, pp. 26- 40 ,(2014) , 10.1109/MRA.2014.2322295
Malavika Bhaskaranand, Jerry D. Gibson, Low-complexity video encoding for UAV reconnaissance and surveillance military communications conference. pp. 1633- 1638 ,(2011) , 10.1109/MILCOM.2011.6127543
Shumeet Baluja, Takeo Kanade, Henry A. Rowley, Human Face Detection in Visual Scenes neural information processing systems. ,vol. 8, pp. 875- 881 ,(1995)
A. Ryan, M. Zennaro, A. Howell, R. Sengupta, J.K. Hedrick, An overview of emerging results in cooperative UAV control conference on decision and control. ,vol. 1, pp. 602- 607 ,(2004) , 10.1109/CDC.2004.1428700
Joshua Gleason, Ara V Nefian, Xavier Bouyssounousse, Terry Fong, George Bebis, None, Vehicle detection from aerial imagery international conference on robotics and automation. pp. 2065- 2070 ,(2011) , 10.1109/ICRA.2011.5979853