Infrared image recognition algorithm based on target character prediction

Aimed at the prevalent phenomenon that it compute too many number and run slowly while infrared imaging air-air missile recognize targets, presented the algorithm that pattern match recognition applied character prediction. Firstly, choose target characters that adapt to recognition of air-air missile. Then the text classified these characters, predicted with weighted average method to the relative character that not changed with time, and predicted with kalman filter to the absolute character that changed rapidly with time. Finally, make the predicted value as the molding board, and used pattern match method to recognize. The result of simulation showed that the algorithm can improve recognition efficiency with the precondition of good recognition purpose.