Abstract: This paper deals with an optimization design method for the Gabor filters based on the analysis of an iris texture model. By means of analyzing the properties of an iris texture image, the energy distributi on regularity of the iris texture image measured by the average power spectrum density is exploited, and the theoretical ranges of the efficient valued frequency and orientation parameters can also be deduced. The analysis shows that the energy distribution of the iris texture is generally centralized around lower frequencies in the spatial frequency domain. Accordingly, an iterative algorithm is designed to optimize the Gabor parameter field. The experimental results indicate the validity of the theory and efficiency of the algorithm.

Key words: texture analysis, iris recognition, receptive profile, Gabor filter, parameter field, optimization method