The key of the star pattern identification algorithm is highly efficient identification with little time.In the past few decades, people committed to methods to search database, and indeed invented a large number of search methods. In order to save the time of searching the star map database, the technology of hierarchical classification of different stars is presented, meanwhile, uses this fast and accurate search method to propose new smart fast star identification algorithm. Desktop-based simulation results show that this star identification methods and database search methods with high accuracy and efficiency. The time complexity of searching star features via database is 0(n). In addition, because the quality of the star image determines accuracy of the star pattern recognition algorithm, a fuzzy edge detection technique is proposed to solve the problem of preprocessing images. This method has a great significance to noise cancellation, the star feature extraction and manufacture and match of database.