Skip to Main Content
Of all the biometric applications available today, it is generally conceded that iris recognition is one of the most accurate. In the past several years a huge amount of iris recognition algorithms have been proposed. However, the vast majority of proposed algorithms restrict to extracting distinct features out of preprocessed iris textures to generate discriminative binary iris-codes, neglecting potential improvements in matching procedures. In this work we present a new technique for matching binary iris-codes. Information of authentication procedures is leveraged by maintaining so-called reliability masks for each user, which indicate local consistency of enrollment templates. Based on user-specific reliability masks a weighted matching procedure is performed in order to improve recognition performance. We apply the proposed matching procedure to different iris recognition algorithms and compare obtained recognition rates to other matching techniques. Experimental results confirm the worthiness of our approach.