Object Detection and Tracking with UAV Data Using Deep Learning
- PDF / 1,344,399 Bytes
- 7 Pages / 595.276 x 790.866 pts Page_size
- 57 Downloads / 412 Views
RESEARCH ARTICLE
Object Detection and Tracking with UAV Data Using Deep Learning A. Ancy Micheal1 • K. Vani1 • S. Sanjeevi2 • Chao-Hung Lin3 Received: 17 October 2020 / Accepted: 22 October 2020 Ó Indian Society of Remote Sensing 2020
Abstract UAVs have been deployed in various object tracking applications such as disaster management, traffic monitoring, wildlife monitoring and crowd management. Recently, various deep learning methodologies have a profound effect on object detection and tracking. Deep learning-based object detectors rely on pre-trained networks. Problems arise when there is a mismatch between the pre-trained network domain and the target domain. UAV images possess different characteristics than images used in pre-trained networks due to camera view variation, altitude ranges and camera motion. In this paper, we propose a novel methodology to detect and track objects from UAV data. A deeply supervised object detector (DSOD) is entirely trained on UAV images. Deep supervision and dense layer-wise connection enriches the learning of DSOD and performs better object detection than pre-trained-based detectors. Long–Short-Term Memory (LSTM) is used for tracking the detected object. LSTM remembers the inputs from the past and predicts the object in the next frame thereby bridging the gap of undetected objects which improves tracking. The proposed methodology is compared with pre-trained-based models and it outperforms. Keywords UAV Deep learning DSOD LSTM Object tracking
Introduction Increasing drone deployment contributes more precise spatial information to the remote sensing community. UAV serves as an intermediate between satellite image data and static camera data; Technical advancement in UAV is increasing rapidly. The up-gradation in navigation systems, sensing payloads, multispectral cameras, orientation systems, laser scanners and thermal imaging has enabled UAV to expand its application in various domains (Colomina et al. 2014). In order to understand the ecosystem, UAVs are deployed to study the tree detection (Selim et al. 2019), monitoring forest recovery from fire and forest restoration study (Zahawi et al. 2015), monitoring the moving patterns
& A. Ancy Micheal [email protected] 1
Department of Information Science and Technology, College of Engineering, Anna University, Chennai 600025, India
2
Department of Geology, College of Engineering, Anna University, Chennai 600025, India
3
Department of Geomatics, National Cheng-Kung University, No.1, University Road, Tainan City 701, Taiwan
of aquatic vertebrates (Raoult et al. 2018), investigating and warning about shark and ray in a coral lagoon (Kiszka et al. 2016), analyzing waterbird population and detecting algae bloom. UAVs are used in disaster management due to their ability to reach the inaccessible areas. In post-disaster management, it can provide an effective response. In military surveillance, identifying and tracking the target with UAV will reduce the threat within the country border (Zaheer et al. 2016). In the vario
Data Loading...
