1 April 2009 Iris recognition with an improved empirical mode decomposition method
Author Affiliations +
Abstract
With the increasing need for security systems, iris recognition is one of the reliable solutions for biometrics-based identification systems. In general, an iris recognition algorithm includes four basic modules: image quality assessment, preprocessing, feature extraction, and matching. This work presents a whole iris recognition system, but particularly focuses on the image quality assessment and proposes an iris recognition scheme with an improved empirical mode decomposition (EMD) method. First, we assess the quality of each image in the input sequence and select clear enough iris images for the succeeding recognition processes. Then, an improved EMD (IEMD), a multiresolution decomposition technique, is applied to the iris pattern extraction. To verify the efficacy of the proposed approach, experiments are conducted on the public and freely available iris images from the CASIA and UBIRIS databases; three different similarity measures are used to evaluate the outcomes. The results show that the presented schemas achieve promising performance by those three measures. Therefore, the proposed method is feasible for iris recognition and IEMD is suitable for iris feature extraction.
© (2009) Society of Photo-Optical Instrumentation Engineers (SPIE)
Jyh-Chian Chang, Ming-Yu Huang, Jen-Chun Lee, Chien-Ping Chang, Te-Ming Tu, "Iris recognition with an improved empirical mode decomposition method," Optical Engineering 48(4), 047007 (1 April 2009). https://doi.org/10.1117/1.3122322 . Submission:
JOURNAL ARTICLE
15 PAGES


SHARE
Back to Top