This paper examines the implementation of high resolution fingerprint sensor chip. The chip is composed of 256/spl times/26 sensing cells. Using this integrated sensor, the fingerprint is captured by pressing the finger skin onto the chip surface. Capacitive sensors that detect the electric field variation induced by the skin surface sample the fingerprint pattern. The automatic fingerprint identification system (AFIS) compares fingerprint based on their differences and similarities of ridge ending and bifurcation. The efficiency is reduced if the database is too large. Partitioning of the fingerprint on the basis of classification of the print into basic patterns known as loop, arch, whirl, and scar is adopted. Once the class of each incoming print has been determined, the set of possible matching print in the database can be restricted thereby reducing the number of comparisons that must be performed.