<![CDATA[Aircraft are widely used in both military and civilian fields today. Detecting aircraft in the airspace is of great strategic and societal importance. In recent years, distinguishing images generated by artificial intelligence from real images has become increasingly difficult. This article presents a study on the classification of real aircraft images and AI-generated aircraft images by machine learning algorithms. Six classifications were obtained from 300 images in the dataset. These classifications are: fake commercial aircraft AI, fake military aircraft AI, fake private aircraft AI, real commercial aircraft, real military aircraft, and real private aircraft. These data were classified using common machine learning models such as Artificial Neural Network (ANN), K-Nearest Neighbor (KNN), Support Vector Machine (SVM), and Logistic Regression (LR). Accuracy, Precision, Recall, and F1 Score metrics were used to analyze the classification success of these models. ROC was used for a detailed analysis of the classification success of the models. According to the results obtained, the ANN model achieved a classification success rate of 96.6%, the KNN model 90.4%, the SVM model 96.7%, and the LR model 96.5%. The highest classification success rate was obtained from the SVM model. These results show that all models achieved similar classification success rates, with the KNN model achieving a lower classification success rate than the others. In conclusion, it can be said that all models can be used in the classification of aircraft images.]]>
Copyrights © 2026
