作者: Chaohong Wu , Sergey Tulyakov , Venu Govindaraju
DOI: 10.1007/11848035_30
关键词:
摘要: Fingerprint image quality is an important factor in the performance of Automatic Identification Systems(AFIS). It used to evaluate system performance, assess enrollment acceptability, and fingerprint sensors. This paper presents a novel methodology for measurement. We propose limited ring-wedge spectral measure estimate global features, inhomogeneity with directional contrast local features. Experimental results demonstrate effectiveness our proposal.