Monitoring infrastructure with remote sensing applications that use object detection, particularly aircraft detection from satellite data, is vital to a country’s military strategy. It ensures complete surveillance and protection of critical assets, enabling proactive measures against potential attacks and boosting national security capabilities. Aircraft detection using deep learning significantly contributes to infrastructure monitoring, thereby strengthening defense decisions. This research compares the performance of object detection models in recognizing aircraft in remote sensing imagery. The models studied included Faster R-CNN, YOLOv8(small, medium, and large variants), YOLO World, and Grounding DINO. These models are fine-tuned using synthetic datasets and very high-resolution (VHR) imagery. All models were fine-tuned independently for each dataset before being tested on an independent VHR image dataset. Faster R-CNN achieved the best performance in both scenarios. When fine-tuned on a synthetic dataset, it reached mAP of 0.7973 and a precision of 0.8248. On a VHR dataset, it achieved a mAP of 0.9326 and a precision of 0.9429.