Background & Purpose: Lately, identification and recognition of objects in an image and video has been attended by many researchers. Aerial images are one of those images supplied by multiple ground-based and space-based tools. Todays, many aerial and space imaging systems are used to provide large amounts of visual data for monitoring and controlling the terrestrial effects.
Methodology: This research is applicable in terms of purpose and the results of this study are presented using simulation - based methods.
Findings: The accuracy of the basic neural network is increased by 7.68% to recognize the objects in aerial images by making some changes in the basic architecture and using two error functions to recognize the objects.
Conclusion: Given that image processing is one of the most reliable automated methods available in defense and surveillance systems to identify and detect objects, increasing the accuracy of these systems will increase the efficiency of these systems.
Zanganeh,A , Sharifi,E and Lotfi,M . (2024). Object recognition in aerial images using deep learning. Journal of Emerging Defense Knowledge & Research Policy, 1(2), 55-70. doi: 10.22034/edrkp.2022.152538
MLA
Zanganeh,A , , Sharifi,E , and Lotfi,M . "Object recognition in aerial images using deep learning", Journal of Emerging Defense Knowledge & Research Policy, 1, 2, 2024, 55-70. doi: 10.22034/edrkp.2022.152538
HARVARD
Zanganeh A, Sharifi E, Lotfi M. (2024). 'Object recognition in aerial images using deep learning', Journal of Emerging Defense Knowledge & Research Policy, 1(2), pp. 55-70. doi: 10.22034/edrkp.2022.152538
CHICAGO
A Zanganeh, E Sharifi and M Lotfi, "Object recognition in aerial images using deep learning," Journal of Emerging Defense Knowledge & Research Policy, 1 2 (2024): 55-70, doi: 10.22034/edrkp.2022.152538
VANCOUVER
Zanganeh A, Sharifi E, Lotfi M. Object recognition in aerial images using deep learning. Emerging Knowledge Policy. 2024;1(2):55-70 (In Persian). doi: 10.22034/edrkp.2022.152538
