Iris image analysis and reconstruction based on zero-crossing wavelet transform

The paper described the iris image analysis and reconstruction procedure based on 1-D zero-crossing wavelet transform. Sharp variation points are among the most meaningful features for characterizing transient signals. As a result, The Sharp variation points can be detected by zero-crossing, then The sharp variation points at a few intermediate levels have been selected as the iris meaningful features for iris recognition. Mallat pointed out that the complete, stable representation based on 1-D zero-crossing wavelet transform under certain restrictive assumptions, In order to compare visually the model to the tester's iris image, the iris model image have to be reconstructed from zero-crossing wavelet transform. Accordingly, we translate 2-D iris image into 1-D signal as boles did. Reconstructing iris image, we translated the 1-D signal into 2-D image. In practice, we used the zero-crossing representation to complete the iris image analysis and reconstruction and the iris recognition system achieved a high efficiency.