| In recent years,the unmanned aerial vehicles have shown a rapid development.More and more UAVs have been widely applied in national defense,aerial photography,agriculture and other areas.At present,there is a lack of effective management on UAVs.The "black fly" incidents were happened frequently around the world,which have endangered the personal privacy and the social security seriously.Therefore,it’s significant to supervise the illegal UAVs with the sound detection and localization methods based on microphone array,and put them into a security system.In this paper,UAV sound is viewed as the target sound source.We research the UAV detection and localization methods based on analysis of the harmonic frequency and the voiceprint features.The main contents of the paper are listed as follows:1.Firstly,we propose a UAV detection method based on the UAV voiceprint features.Based on the analysis of the sound spectrum characteristics of UAV,we select MFCC(Mel-scale Frequency Cepstral Coefficients)feature vector,the voiceprint energy feature vector,and the combination of them in order to achieve good discrimination of the UAV sound.Then,we use SVM(Support Vector Machine)method to classify sounds into two categories:the AUV sound or other sounds by which the drones can be detected.The experimental results show that the proposed method has achieved good detection results.The detection rate is above 80%within 80m in the open area of the city.And the false alarm rate is less than 15%.2.Secondly,we propose a localization method based on time-delay estimation.The time-delay is calculated by the generalized cross-correlation(GCC)algorithm with the weighting function of the phase transform(PHAT).With two L-shaped microphone arrays,we can obtain the three-dimensional positions of the UAV using the time-delay calculated by the GCC.3.Finally,we establish an experimental system for detection and localization of a UAV.The detection and localization system consists of two microphone arrays,a data acquisition module,a data transmission module and algorithms based on the LabVIEW.The experimental results have shown that we have realized real-time detection and localization of UAV. |
