Abstarct: The design and production of voice recognition systems is the research goal of many scientific centers in the last half-century. One of the human goals in the production of such systems is of course the fact that the entry of information and the implementation of instructions in audio in addition to saving time and cost in different ways also increases the quality of life. The subject matter of this thesis is to provide an intelligent vehicle identification and classification algorithm with its voice processing which is now very important for reasons of use in military and civilian systems. The proposed algorithm should be able to use the car's voice identify and classify its type. The first step in implementing this algorithm is to collect the databaxse. In this thesis, the databaxse has been compiled from the 6 classes of BMW 540i, Ferrari 308, Ford Tempo, Jeep, Police and Volvo Amazon each of which contains 51 samples. The performance of the algorithm is to extract the four important parameters of the MFCC, LPC, LPCC and LSF with the length of the characteristics of 13, 26, and 39 coefficients per frxame of the recorded voice of the vehicles and as corresponding vehicle reference features Give to the SVM classifier to be trained with these features. In order to evaluate the algorithm the system extracts the characteristics of an audio as an input and give it into the SVM classifier then classifies the type of vehicle by comparing the input sound characteristics with the reference characteristics. The results of the implementation of this algorithm show that the LSF technique with the detection accuracy of 85.02% of the best feature extraction technique and the MFCC method is also with accuracy of 82.94% it is ranked second. On the other hand the feature extraction techniques of the LPC and LPCC respectively with accuracy of 55.62% and 58.9% for this Identification algorithm can not be good results and their use is not useful.