We present a privacy preserving fingerprint authentication protocol based on minutiae points which is the most accurate and promising fingerprint matching algorithm. Our protocol provides a secure alignment and matching protocol in two-party settings between a user and a service provider using inner product encryption. We devise a new encodings to compute alignment parameters consists of shift and rotation information for minutiae points. This allows the user to align fingerprint before encrypting fingerprint sample. The implementation results over desktop takes less than 1 second in authenticating one person which shows practicality of our protocol. This results achieve at least 60 times faster than previous 2-party setting of recent work which exploits Garbled circuit.