A new approach for noisy iris database indexing based on color information

Iris indexing is one of the hot topics in the research on iris recognition. Effective indexing method will both reduce the computational complexity and enhance the matching accuracy of iris recognition systems. In this paper, a new noisy iris indexing approach based on iris color information is proposed. After color correction, this method utilizes normalized color components to index the noisy iris images. Hit rate and penetration rate are used to measure the indexing and retrieval performance of the proposed method. The effectiveness evaluation is performed based on the UBIRIS.v2 database, and the results are analyzed and discussed.