Fingerprints are commonly utilized as a key technique and for personal recognition and in identification systems for
personal security affairs. The most widely used fingerprint systems utilizing the distribution of minutiae points for fingerprint
matching and representation. These techniques become unsuccessful when partial fingerprint images are capture, or the finger ridges
suffer from lot of cuts or injuries or skin sickness. This paper suggests a fingerprint recognition technique which utilizes the local
features for fingerprint representation and matching. The adopted local features have determined using Haar wavelet subbands. The
system was tested experimentally using FVC2004 databases, which consists of four datasets; each set holds 80 fingers with 8 samples
per finger. The attained test results showed encouraging system capability to recognize low quality fingerprint images although with
presence of partial loss in fingerprint images.