In this paper, classification of hand gestures for the smart control of prosthetic hands is proposed. The surface Electromyography (sEMG) signals are used for classifying the hand gestures. The important attributes of the signal are extracted by finding Hilbert Huang Transform (HHT). These features are given as input to the Deep Neural Network (DNN) classifier for further classification. The experimental results show that high classification accuracy can be achieved for the proposed method compared to the other techniques.