Dear Buhari.

I recommend to you for first to separate the process, in minutiae extraction I suggest use for pre processing the STFT strategie, to binarize the resultant image the most important thing is how you select the treshold and you can use this paper to have an idea: Adaptive Fingerprint Binarization by Frequency Domain Analysis, to the thinning process you can use this paper A Modified fingerprint image thinning algorithm, the cross number is the algorithm user to extract the minutiae point and then select a post processing strategie in order to eliminate the spurius minutiae from the extraction process. For comparision I suggest MCC. Those are the main process to implement the system but there are a lot of more.

Best regards

Several researchers have done a comparative analysis of finger printing algorithms such as:

http://pubs.sciepub.com/jcsa/1/4/1/.

http://www.cse.nd.edu/~mblanton/papers/esorics11.pdf

https://www.utdallas.edu/~bxt043000/Publications/Conference-Papers/DAS/C77_Fingerprint_Matching_Algorithm.pdf

Take a look here because this is a quasi-standard which is usually cited:

http://www.fbi.gov/about-us/cjis/fingerprints_biometrics/biometric-center-of-excellence/files/fingerprint-recognition.pdf