Iris Recognition Using Fuzzy Support Vector Machine With Several Kernels
Agus Kurniawan · IPB Scientific Repository (Bogor Agricultural University.) · 2008
The authentication of people using iris-based recognition is a widely developing technology. Iris recognition can be used in differentiating identical twins. It is possible because iris has a unique structure. In this research, an iris recognition system using Fuzzy Support Vector Machine (FSVM) was proposed. Using the decision functions obtained by training the FSVM, for each class, a truncated polyhedral pyramidal membership function could be defined. The performance of the proposed iris recognition system was analyzed using CASIA database 3.0. The original images were non-ideal and thus required a sequence of pre-processing steps prior to application of encoding method. Experimental results show that the iris recognition method proposed has a recognition rate about 90.75%.