Privacy-Preserving Federated Learning System (f-PPLS) for military focused area classification

Accurate classification of military-focused areas using machine learning techniques is crucial for meeting military criteria. However, preserving high data privacy in aerial image classification poses significant challenges. This paper proposes a Privacy-Preserving Federated Learning System (f-PPLS) for military-focused area classification. The aim of this paper is to provide data privacy preserved aerial image classification technique for identification of military area images using federated learning mechanism. Here, federated learning have been applied to achieve data privacy on Identically distributed (IID) and non-identically distributed (non-IID) military area aerial images. By leveraging federated learning, the system enables collaborative learning across multiple military organizations without sharing raw data, ensuring data privacy and security. The system utilizes local datasets from participating organizations, containing information about areas such as runways, storage tanks, harbors, coasts, buildings, and airplanes. The system achieves a comparable classification accuracy to traditional approaches, as evidenced by numerical results indicating 100% accuracy. Experimental results on real-world military-focused area datasets demonstrate the effectiveness and efficiency of the f-PPLS system, achieving comparable classification accuracy to traditional approaches while preserving data privacy and security. By utilizing collaborative deep learning techniques and ensuring data privacy, the f-PPLS system offers a practical and effective solution for accurate military-focused area classification.