This paper investigates the current methods for fingerprint template recognition and database security. This target is mainly decomposed into image preprocessing, minutiae extraction, minutiae matching and template security. This paper proposes a two level transformation and encryption techniques for fingerprint minutiae matching to reduce error rate as well as reduce the computational effort from the existing method. For each sub-task, some classical and up-to-date methods in literature are analyzed. Based on the analysis integrated solutions for fingerprint recognition and template security is developed for demonstration. Demonstration program is coded by OPENCV. For the program some optimization at code implementation level and algorithm level are proposed to increase the result to 95%. These performance enhancements are shown by experiments conducted upon a variety of fingerprint images. Also the experiments illustrate the key issues of fingerprint security that are consistent with what the available literatures say. © 2014 IEEE.