A new recognition algorithm with high result reliability

High recognition rate is the aim of pattern recognition, which is difficult to obtain without ideal feature extraction. To solve this problem, a new recognition algorithm with high result reliability is proposed. The core contents of this algorithm are MSCM (multiple-set-compete method) and the algorithm of result reliability. In this paper, MSCM improving the recognition rate with nonideal feature extraction and the algorithm of result reliability are illustrated. A specific recognition case of 16 classes of sounds and the test flow of the specific case are given. The results show that the new recognition algorithm can improve the recognition rate greatly.