We present here the practical implementation of a biometric system for recognition of persons based on the PM demodulation of the iris image, through the use of a self-similar, multi-scale Wavelet family of two-dimensional frequency-selective quadrature filters that use Gabor basic functions. The identity verification is based on the execution of a test of statistical independence between the resulting complex codes of the image projection on the filters, by means of the measure of the normalized Hamming distance and establishing a criterion as of a strategy curve for the decision. The accomplishment of this project this mainly based on the results that fearlessness the pioneer work of Dr. John G. Daugman [1] on this particular biometric technique in 1993. The framework of the application is the new technology of the Virtual Instrumentation, using a conventional video camera, an image acquisition board and a personal computer with the use of IMAQVision and LabView6i as a development platform.