Autonomous aviation combat strategist using AI and computer vision

With the growing complexity of modern aerial warfare and the increasing need for autonomous decision-making, the integration of Artificial Intelligence (AI) and Computer Vision into air combat strategies has become indispensable. This study presents an advanced aircraft detection and classification framework using a customized YOLOv8-based deep learning architecture, optimized for high-resolution aerial and satellite imagery. The proposed model addresses key challenges such as sub-pixel target recognition, dense object clutter, and class confusion under varied atmospheric and lighting conditions. Comparative evaluation against the baseline YOLOv5m model demonstrates substantial improvements in key performance metrics, including a 5.7% gain in mAP@0.5 and a 5.45% increase in F1-Score. Furthermore, the model achieves a 13.4% improvement in detecting small objects, with a reduced parameter count and enhanced inference speed making it well-suited for real-time edge deployment in autonomous aerial systems. Experimental results affirm the model's robustness, accuracy, and potential applicability in real-time tactical surveillance, unmanned combat aerial vehicles (UCAVs), and defense intelligence operations. This work contributes a scalable and efficient solution for future-ready, AI-driven air combat ecosystems.