Abstract

W artykule opisano aplikację, która automatycznie wykrywa zdarzenia dźwiękowe takie jak: rozbita szyba, wystrzał, wybuch i krzyk. Opisany system składa się z bloku parametryzacji i klasyfikatora. W artykule dokonano porównania parametrów dedykowanych dla tego zastosowania oraz standardowych deskryptorów MPEG-7. Porównano też dwa klasyfikatory: Jeden oparty o Percetron (sieci neuronowe) i drugi oparty o Maszynę wektorów wspierających. Dokonano porównań wyników i przedyskutowano możliwości praktycznego zastosowania systemu. Digital signal processing of the sound is a domain with numerous applications in the telecommunications and informatics. These well developed algorithms of the analysis of the sound can be also applied in the field of security systems, where the traditional monitoring is still based mainly on video cameras. The commonly used monitoring cameras can be equipped with additional microphones and the audio content can be analyzed by a monitoring program running on a dedicated hardware. This application can automatically detect in the audio stream events like a broken window, gun-shot, explosion or scream. One of the main parts of this system is a parameterization block. In the paper two parameterization methods are proposed for this purpose. First of them is based on the frequency analysis of the examples of the sound events. Second one is based on using a standardized set of audio MPEG-7 and cepstral descriptors. The feature vectors calculated by these two methods have been used for the training of two intelligent classifiers: a Support Vector Machines classifier (SVM) and a Neural Networks Perceptron (NNP). The classifiers have been verified by using of the cross validation method. The results have been compared and conclusion derived. The application of the results in a system working in real conditions is presented and discussed at the end of the paper. The work has been done in the frame of international project: "INDECT" (Intelligent Information System Supporting Observation, Searching and Detection for Security of Citizens in Urban Environment).

Citations

Authors (2)