Title :
Fingerprint image segmentation based on Fingerprint Ridge Intensity
Author :
Tong, Xi-feng ; Li, Peng-fei
Author_Institution :
Sch. of Comput. & Inf. Technol., Northeast Pet. Univ., Daqing, China
Abstract :
A fingerprint image segmentation approach is proposed in this paper. First, a feature called Fingerprint Ridge Intensity (FRI) is described in detail, which employs two intrinsic fingerprint features, that is, ridge orientation and ridge circle, to represent ridge intensity. Then an adaptive segmentation algorithm for FRI is proposed. Finally, a post-processing procedure is proposed to reduce segmentation noise. Evaluation on FVC2004 fingerprint databases reveal that the proposed approach is good at segmenting fingerprint image.
Keywords :
fingerprint identification; image segmentation; visual databases; FRI; FVC2004 fingerprint database; adaptive segmentation algorithm; fingerprint image segmentation; fingerprint ridge intensity; noise segmentation; ridge circle; ridge intensity; ridge orientation; Cybernetics; Databases; Equations; Fingerprint recognition; Image matching; Image segmentation; Machine learning; FVC2004; Fingerprint; Fingerprint Ridge Intensity; Image segmentation;
Conference_Titel :
Machine Learning and Cybernetics (ICMLC), 2011 International Conference on
Conference_Location :
Guilin
Print_ISBN :
978-1-4577-0305-8
DOI :
10.1109/ICMLC.2011.6016981
