Iris recognition method based on ordinal measure of discrete cosine transform coefficients

A recognition system based on irides has become important in the last decades due to its reliability and comfort. This paper proposed a method for iris recognition based on ordinal measure of Discrete Cosine Transform (DCT) coefficients. Ordinal measure was obtained by ordering the absolute value of AC coefficients of normalized iris image of both database and query image ascendantly. Distance between those image was measured by Minkowsky distance metrics. Five simulation sets, each with different of AC coefficient used, namely all, 63, 48, 32 and 16 coefficients were conducted, to find out recognition rate of the proposed method and the best trade-off between compression ratio and recognition rate. It turned out that 50% of querying task resulted in averaged recognition rate more than 60%, Moreover, the best trade-off between compression ratio and recognition rate was achieved when as many as 48 coefficients were used.