scholarly journals Deep Learning-Based Ground Target Detection and Tracking for Aerial Photography from UAVs

2021 ◽  
Vol 11 (18) ◽  
pp. 8434
Author(s):  
Kaipeng Wang ◽  
Zhijun Meng ◽  
Zhe Wu
Keyword(s):  
Deep Learning ◽  
Target Tracking ◽  
Target Detection ◽  
Aerial Photography ◽  
Tracking Algorithm ◽  
Aerial Image ◽  
Small Scale ◽  
Detection And Tracking ◽  
Tracking Algorithms ◽  
Ground Target

Target detection and tracking can be widely used in military and civilian scenarios. Unmanned aerial vehicles (UAVs) have high maneuverability and strong concealment, thus they are very suitable for using as a platform for ground target detection and tracking. Most of the existing target detection and tracking algorithms are aimed at conventional targets. Because of the small scale and the incomplete details of the targets in the aerial image, it is difficult to apply the conventional algorithms to aerial photography from UAVs. This paper proposes a ground target image detection and tracking algorithm applied to UAVs using a revised deep learning technology. Aiming at the characteristics of ground targets in aerial images, target detection algorithms and target tracking algorithms are improved. The target detection algorithm is improved to detect small targets on the ground. The target tracking algorithm is designed to recover the target after the target is lost. The target detection and tracking algorithm is verified on the aerial dataset.

UAV Aerial Photography Target Detection and Tracking Based on Deep Learning

2021 ◽  
pp. 426-438
Author(s):  
Xiaohua Li ◽  
Feiyang Wang ◽  
Aiming Xu ◽  
Guofeng Zhang
Keyword(s):  
Deep Learning ◽  
Target Detection ◽  
Aerial Photography ◽  
Detection And Tracking

GROUND TARGET DETECTION AND TRACKING

2001 ◽  
pp. 171-176
Author(s):  
PHILIPPE LACOMME ◽  
JEAN-PHILIPPE HARDANGE ◽  
JEAN-CLAUDE MARCHAIS ◽  
ERIC NORMANT
Keyword(s):  
Target Detection ◽  
Detection And Tracking ◽  
Ground Target

HE-YOLO: Aerial Target Detection Based On Improved YOLOv3

2021 ◽  
Author(s):  
Yuqing Zhao ◽  
Jinlu Jia ◽  
Di Liu ◽  
Yurong Qian
Keyword(s):  
Target Detection ◽  
Loss Function ◽  
Detection Rate ◽  
Aerial Image ◽  
Experimental Comparison ◽  
Small Scale ◽  
Detection Accuracy ◽  
False Detection Rate ◽  
False Detection ◽  
Bounding Box

Aerial image-based target detection has problems such as low accuracy in multiscale target detection situations, slow detection speed, missed targets and falsely detected targets. To solve this problem, this paper proposes a detection algorithm based on the improved You Only Look Once (YOLO)v3 network architecture from the perspective of model efficiency and applies it to multiscale image-based target detection. First, the K-means clustering algorithm is used to cluster an aerial dataset and optimize the anchor frame parameters of the network to improve the effectiveness of target detection. Second, the feature extraction method of the algorithm is improved, and a feature fusion method is used to establish a multiscale (large-, medium-, and small-scale) prediction layer, which mitigates the problem of small target information loss in deep networks and improves the detection accuracy of the algorithm. Finally, label regularization processing is performed on the predicted value, the generalized intersection over union (GIoU) is used as the bounding box regression loss function, and the focal loss function is integrated into the bounding box confidence loss function, which not only improves the target detection accuracy but also effectively reduces the false detection rate and missed target rate of the algorithm. An experimental comparison on the RSOD and NWPU VHR-10 aerial datasets shows that the detection effect of high-efficiency YOLO (HE-YOLO) is significantly improved compared with that of YOLOv3, and the average detection accuracies are increased by 8.92% and 7.79% on the two datasets, respectively. The algorithm not only shows better detection performance for multiscale targets but also reduces the missed target rate and false detection rate and has good robustness and generalizability.

scholarly journals A split target detection and tracking algorithm for ballistic missile tracking during the re-entry phase

2020 ◽  
Vol 16 (6) ◽  
pp. 1142-1150
Author(s):  
Muhammad Asad ◽  
Sumair Khan ◽  
Ihsanullah ◽  
Zahid Mehmood ◽  
Yifang Shi ◽  
...  
Keyword(s):  
Target Detection ◽  
Ballistic Missile ◽  
Tracking Algorithm ◽  
Detection And Tracking

An Anti- Occlusion Tracking Algorithm Based on MCD Correlation Matching

2013 ◽  
Vol 718-720 ◽  
pp. 2005-2010
Author(s):  
Pu Liu ◽  
Chun Ping Wang ◽  
Qiang Fu
Keyword(s):  
Target Tracking ◽  
Template Matching ◽  
Tracking Algorithm ◽  
Experimental Results ◽  
Template Method ◽  
Target Movement ◽  
Movement Prediction ◽  
Tracking Algorithms ◽  
The Stability ◽  
Correlation Matching

In order to improve the stability of target tracking under occlusion conditions，on the basis of researching some target tracking algorithms, this paper presents an algorithm based on MCD correlation matching, which combines multi sub-templates matching and target movement prediction. Besides, for occlusion characteristics, corresponding template matching criterions and updating methods are put forward. Experimental results show that, comparing with the single template method which updating frame by frame, the proposed algorithm has a certain anti-occlusion ability with better stability and continuity of target tracking under occlusion conditions.

scholarly journals Deep Learning for Component Fault Detection in Electricity Transmission Lines.

2021 ◽  
Author(s):  
Iyke Maduako ◽  
Chukwuemeka Fortune Igwe ◽  
James Edebo Abah ◽  
Obianuju Esther Onwuasoanya ◽  
Grace Amarachi Chukwu ◽  
...  
Keyword(s):  
Deep Learning ◽  
Transmission Lines ◽  
Ground Truth ◽  
Recall Rate ◽  
Aerial Image ◽  
Small Scale ◽  
Single Shot ◽  
Deep Convolutional Neural Networks ◽  
Electricity Transmission ◽  
Human Cost

Abstract Fault identification is one of the most significant bottlenecks faced by electricity transmission and distribution utilities in developing countries to deliver efficient services to the customers and ensure proper asset audit and management for network optimization and load forecasting. This is due to data scarcity, asset inaccessibility and insecurity, ground-surveys complexity, untimeliness, and general human cost. In view of this, we exploited the use of oblique UAV imagery with a high spatial resolution and a fine-tuned and deep Convolutional Neural Networks (CNNs) to monitor four major Electric power transmission network (EPTN) components. This study explored the capability of the Single Shot Multibox Detector (SSD), a one-stage object detection model on the electric transmission power line imagery to localize, detect and classify faults. The fault considered in this study include the broken insulator plate, missing insulator plate, missing knob, and rusty clamp. Our adapted neural network is a CNN based on a multiscale layer feature pyramid network (FPN) using aerial image patches and ground truth to localise and detect faults via a one-phase procedure. The SSD Rest50 architecture variation performed the best with a mean Average Precision (mAP) of 89.61%. All the developed SSD based models achieve a high precision rate and low recall rate in detecting the faulty components, thus achieving acceptable balance levels of F1-score and representation. Finally, comparable to other works in literature within this same domain, deep-learning will boost timeliness of EPTN inspection and their component fault mapping in the long - run if these deep learning architectures are widely understood, adequate training samples exist to represent multiple fault characteristics; and the effects of augmenting available datasets, balancing intra-class heterogeneity, and small-scale datasets are clearly understood.

Sign in / Sign up

Export Citation Format

Share Document