Biometric authentication plays a major role in security as these are by nature unique for every human. But the security is compromised when the pattern matching system is not accurate. Authentication system like fingerprint recognition is most commonly used biometric authentication system. In this paper survey is done on fingerprint recognition techniques. And different approaches are studied in terms of accuracy and performance. As fingerprint may also contain noise; so image de-noising techniques are also studied Cross ridge frequency analysis of fingerprint images is performed by means of statistical measures and weighted mean phase is calculated. These different features along with ridge reliability or ridge centre frequency are given as inputs to a fuzzy c-means classifier..

Biometric verification assumes a noteworthy part in security as these are by nature interesting for each human. Be that as it may, the security is traded off when the example coordinating framework is not exact. Verification framework like unique finger impression acknowledgment is most ordinarily utilized biometric validation framework. In this paper overview is done on unique mark acknowledgment strategies. What's more, extraordinary methodologies are examined as far as precision and execution. As unique finger impression may likewise contain clamor; so picture de-noising systems are additionally concentrated Cross edge recurrence examination of finger impression pictures is performed by method for factual measures and weighted mean stage is ascertained. These distinctive components alongside edge unwavering quality or edge focus recurrence are given as contributions to a fluffy c-implies classifier.