Performance Analysis on the Assessment of Fingerprint Image Based on Blur Measurement

Abstract

Processing the fingerprint image captured from the mobile phone camera is problematic. The addressed problem is such as degree of freedom, blurry, noisy and weak ridges-valleys images. In this paper, problem of blur assessment on fingerprint image is the main consent. In this regards, dataset of fingerprint image is collected from several subjects and different environment to obtain a clear and blur fingerprint image. The ground truth for this dataset is assigned by human vision. According to literature, the suitable blur evaluation is based on no-reference strategy. Hence, this paper modified existing no-reference blur assessment so it is suitable for fingerprint image. The proposed method achieved promising result and able to detect blur and clear fingerprint image.