Analysis and Comparison of Image-Based UAV Detection and Identification

This paper presents Unmanned Aerial Vehicles (UAVs) detection and classification with the help of different image-based machine learning modalities. The field of UAVs attracted researchers in recent years in response to the exponential rise in the number of UAVs available in the market with applications ranging from entertainment to defense operations and the risk associated risk by the same. Presently, visual, radar, radio frequency, and acoustic sensing systems are the prevailing technologies in the field of detection and identification of UAVs. The general results of this study show that UAV machine learning-based classifications are propitious and that there are many successful individual contributions. In this research, UAVs were detected and classified using classification methods like Support Vector Machines (SVM), k-nearest neighbor (KNN), and Convolutional Neural Networks (CNN). The results demonstrated that CNN, SVM, and KNN had an accuracy of 91%, 87%, and 78%, respectively. The classifier CNN outperformed other classifiers under the same empirical circumstances.