Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Feb 2;22(3):1147.
doi: 10.3390/s22031147.

Automatic Target Detection from Satellite Imagery Using Machine Learning

Arsalan Tahir  1 , Hafiz Suliman Munawar  2 , Junaid Akram  3   4 , Muhammad Adil  1 , Shehryar Ali  5 , Abbas Z Kouzani  5 , M A Pervez Mahmud  5
Affiliations

Automatic Target Detection from Satellite Imagery Using Machine Learning

Arsalan Tahir et al. Sensors (Basel). .

Abstract

Object detection is a vital step in satellite imagery-based computer vision applications such as precision agriculture, urban planning and defense applications. In satellite imagery, object detection is a very complicated task due to various reasons including low pixel resolution of objects and detection of small objects in the large scale (a single satellite image taken by Digital Globe comprises over 240 million pixels) satellite images. Object detection in satellite images has many challenges such as class variations, multiple objects pose, high variance in object size, illumination and a dense background. This study aims to compare the performance of existing deep learning algorithms for object detection in satellite imagery. We created the dataset of satellite imagery to perform object detection using convolutional neural network-based frameworks such as faster RCNN (faster region-based convolutional neural network), YOLO (you only look once), SSD (single-shot detector) and SIMRDWN (satellite imagery multiscale rapid detection with windowed networks). In addition to that, we also performed an analysis of these approaches in terms of accuracy and speed using the developed dataset of satellite imagery. The results showed that SIMRDWN has an accuracy of 97% on high-resolution images, while Faster RCNN has an accuracy of 95.31% on the standard resolution (1000 × 600). YOLOv3 has an accuracy of 94.20% on standard resolution (416 × 416) while on the other hand SSD has an accuracy of 84.61% on standard resolution (300 × 300). When it comes to speed and efficiency, YOLO is the obvious leader. In real-time surveillance, SIMRDWN fails. When YOLO takes 170 to 190 milliseconds to perform a task, SIMRDWN takes 5 to 103 milliseconds.

Keywords: SIMRDWN; SSD; YOLO; deep learning; faster RCNN; satellite images.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
Major contributions of object detection frameworks and convolutional neural networks.
Figure 2
Figure 2
This satellite image is taken by Google Earth. It is an open-source platform for collection of satellite images. This image contains objects of interest, i.e., airplanes.
Figure 3
Figure 3
Pipeline of object detection implemented in this research. It starts from data collection followed by model training and making predictions.
Figure 4
Figure 4
Overview of Faster RCNN network layers.
Figure 5
Figure 5
YOLO model divides image into S × S grid and calculates confidence scores with B bounding boxes and predictions are enclosed into (S × S) × (B *5 + C) tensor.
Figure 6
Figure 6
SSD uses feature layers at the end of base network, which predicts the class score with four offset values to default boxes.
Figure 7
Figure 7
Loss after training of model faster RCNN.
Figure 8
Figure 8
Loss function after training the YOLO Model.
Figure 9
Figure 9
Loss function after training the SSD Model.
Figure 10
Figure 10
Loss function after training SIMRDWN Model.
Figure 11
Figure 11
Detection result comparison of faster RCNN, SSD and YOLO.
Figure 12
Figure 12
Detection results of SIMRDWN on1920 × 1080 resolution satellite image.
Figure 13
Figure 13
Detection results of SIMRDWN on a 6088 × 8316 resolution satellite image.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Everingham M., Van Gool L., Williams C.K.I., Winn J., Zisserman A. The pascal visual object classes (VOC) challenge. Int. J. Comput. Vis. 2010;88:303–338. doi: 10.1007/s11263-009-0275-4. - DOI
    1. Ren S., He K., Girshick R., Sun J. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. IEEE Trans. Pattern Anal. Mach. Intell. 2017;39:1137–1149. doi: 10.1109/TPAMI.2016.2577031. - DOI - PubMed
    1. Munawar H.S., Mojtahedi M., Hammad A.W.A., Kouzani A., Mahmud M.A.P. Disruptive technologies as a solution for disaster risk management: A review. Sci. Total Environ. 2021:151351. doi: 10.1016/j.scitotenv.2021.151351. - DOI - PubMed
    1. Sherrah J. Fully Convolutional Networks for Dense Semantic Labelling of High-Resolution Aerial Imagery. arXiv. 2016 preprint .1606.02585
    1. Liu W., Anguelov D., Erhan D., Szegedy C., Reed S., Fu C.-Y., Berg A.C. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Springer; Berlin/Heidelberg, Germany: 2016. SSD: Single Shot MultiBox Detector; pp. 21–37. - DOI

LinkOut - more resources