In this paper, an automated high-accuracy acoustic-based drone detection and classification system using a frequency feature extraction method based on the Prony algorithm is proposed. To show the efficiency of the proposed method, the accuracy of detection and classification of the suggested method is evaluated by an experimental dataset composed of recorded drone audio in a natural environment and under different conditions. Results show more than 97.7% and 93.6% accuracy for detection and classification, respectively, while using typical audio features like MFCCs, GTCCs, and FFT provide less than 78.6% accuracy for classification.