Mobile object recognition based on acoustic information

The paper considers a statistical method to classify objects based on sound emitted by the objects. Engines and the carriageable devices are the sources of the signal. The methodology used doesn't require detailed analysis of which part of the object is responsible for the components of the signal. The sound is preprocessed using DSP methods to obtain feature vectors to be used by a statistical classifier. A computer system equipped with a DSP processor for mobile object recognition as real-time application is also presented. Results of classification of various military vehicles were the subject of conclusions.