This material is available only on bibliotequer
13 Nov 2021 
Claude 
In the ciliary part of the iris in the diagram shows that the dots indicate circular contaction sillones and the papillary part (A), on eachside of the collar there are what are called the Fuchs crypts (C)which are from the root of the iris and also near the pupillary and ciliary part, at the border of the pupil we find the pigmented collar. The surface of the iris is blue in color is dense month for the anterior border and trabeculae moresalient.
The region of the pupil is described by veins, arteries and radial branches, the incomplete arterial minor circle (F) isformed by the arteries and also capillary arches are sensed byso-called branches that extend towards the pupil. The layer (G) describes the circular arrangement of the sphincter muscle, there is also the layer (H) which describes the radial processesof the dilator muscle. The part (I) and (J) which are respectively the radial contraction grooves and the Structural folds of Schwalbe form the upper layer of the iris and finally the last layer (K) is called plicata of the ciliary body [6].
An iris recognition system consists of several stages. The first step is the image acquisition phase, which captures an Fig. 5. Architecture of iris recognition system with a specific camera, which can isolate external lightand the noise effect of the individual. All this reduces thecalculation error and takes a good image quality. The second phase is the segmentation, which consists of the localization ofthe pupil and the iris. Among the methods that are popular, the integrator-differential operator method of John G Daugman [7], he is a pioneer in his field.
Another method is to transform itfrom circular hough to wildes and al [8]. The normalization phase consists in transforming the iris image information into arectangle [9]. The feature extraction step consists in presentingthe normalized
Several techniques of iris segmentation exist in the literature, we find John Daugmans Integro-Diffential Operatormethod [7], Richard Paul Wildes Hough Transform method[8], Libor Maseks method for a recognition [11], also Proençaand Alexander propose a method based on Fuzzy Classification Algorithm [12] He, Tan, Sun, and Qiu propose Pulling and Pushing (PP) [13], Tan, He, and Sun propose amethod based on the grouping of eight neighbors based on the connection [14].
The method of John Daugman is the most used and the most cited in the literature, it is implemented in many iris recognition systems. The principle of the method consists inusing an integro-differential operator for the localization ofthe iris and the pupil with two circles as well as the arches ofthe upper and lower eyelid. The integro-differential operatoris defined as follows [7]: The operator searches for the circular path where there isa maximum change in pixel values, by varying the radius and the center position x, y of circular outline. The operator is applied iteratively to the amount of smoothing which decrease progressively to a precise location [7].
Another important point in Daugmans method is the threshold function, which is characterized by the minimum and maximum values of the iris radius which are called rmin and rmax [29]. The values of the iris radius must be manually set as input before testing and it depends on the type of iris database and the resolution size of the iris image.