Target Recognition in SAR Images for Military Applications

DOI

No DOI

Published In

Volume 5, Issue 2 (2026)

Published Online

May-August 2026

Pages

81-87

Article Type

Original Article

Check for updates

No DOI available for cross-reference

Timely and precise intelligence sits at the heart of sound military decision-making. This paper presents a web-based Defense Intelligence and Asset Management System that pairs a SAR-focused target recognition pipeline with practical security and reporting features. General users can browse de-fense news freely, while administrator access is locked behind Gmail-integrated two-factor authentication that sends a one-time password directly to a registered email account before granting entry. Privileged users can upload multiple SAR images for batch analysis, manage military vehicle records through full CRUD operations, and download structured PDF reports that capture detection results, confidence scores, and auto-generated tactical notes for each identified target. At the core of the analysis layer, a YOLO model locates and classifies tanks, trucks, and ships within SAR frames. Because labeled SAR data is genuinely scarce, two generative networks — DH-GAN synthesize additional training samples that mirror authentic radar backscatter and speckle behavior, keeping the detector accurate under noisy or unusual conditions. Detected targets are logged automatically with spatial coordinates, timestamps, and class information, and the system produces short tactical assessments to assist operators in evalu-ating what each finding means for the mission at hand. Taken together, the components form a self-contained tool that handles the full path from raw SAR imagery to a shareable intelligence report.