Ship classification by sound signature

Nowadays, either for supervision or for defense purposes, the transportation is being controlled by RADAR (Radio Detecting and Ranging). RADAR helps us get information about objects of interest by transmitting electromagnetic radio waves to the target and analyzing the reflecting signals. We understand that in order to determine any given object´s information (speed, distance etc) the first action is always taken by the control center. This also tells us that the moving object can also notice the electromagnetic radio waves and hence the control center can also be sensed. It is of great importance in the defense systems that the object under inspection should not notice being recognized. If we could get information about the target by controlling the waves sent by the target rather than sending any signals to it we could be on the advantageous side. In this study the vehicles were classified by the sounds they make while operating. Working units of the marine vehicles (engine, fan, the generator and etc) analyzed and the sounds they diffuse to the environment while operating were characterized. In the characteristics determination the MFCC (Mel Frequency Cepstral Coefficients) method used to determine the features of the vehicles and by the k-means method, the defined features degraded to a specific number, k-NN (k-Nearest Neighbor) method selected for the classification and recognition of the vehicles.