Indiana University-Purdue University Indianapolis (IUPUI) / Externally visible characteristics (EVCs) predicted from an unknown sample of DNA are particularly useful in forensics as they can provide information beyond that of an STR profile. Current EVCs which are highly studied and well-predicted include iris, hair, and skin color. Notably, models predicting iris color, such as IrisPlex, are the most accurate with up to ~95% accuracy; however, some inaccurate predictions occur, as is evidenced by the ~5%. Often, these are due to green or hazel eyes, which are frequently viewed as intermediate. Though, some of the inaccurate predictions are due to true-blue being predicted as brown and vice versa. Previous research has theorized the possibility of two SNPs, rs12913832 and rs1800407, acting as a functional haplotype affecting iris color. rs12913832 is recognized as the most predictive SNP for iris color and highly significant in other pigmentation phenotypes; presently, rs1800407 is the second-ranked SNP in the IrisPlex 6-SNP system. Both SNPs are highly variable in Europe, where the majority of variation in iris color originates. In the present study, we explore the SNP variation present in the genetic regions of OCA2-HERC2 as well as possible haplotypes. Our research centers around the functional haplotype and the addition of SNPs to the functional haplotype. In addition, three different ways of classifying the phenotype are assessed simultaneously. First, using a 4-point categorical phenotype—blue/blue grey, blue/green yellow, hazel/light brown, and dark brown. Second, calculating a continuous scale from a quantitative phenotype in which the percentage of each categorical color has been measured. Third, using the IrisPlex 6-SNP system to predict eye color and identify individuals which have been inaccurately predicted. Exploration of the SNP and haplotype variation resulted in two SNPs for both the categorical and quantitative phenotypes which were significantly correlated with hazel/light brown—rs1448484 and rs61335644, both as independent SNPs and when assessed in a haplotype with rs1800407-rs12913832. SNP rs1448484 has been associated with skin pigmentation previously and is located in a possible transcription factor binding site. SNP rs61335644 is not presently associated with pigmentation but is in complete LD with two SNPs in and around regulatory regions present in HERC2. Finally, the addition of rs1448484 and rs61335644 into the current IrisPlex 6-SNP system slightly improved each of the tested performance metrics for hazel/light brown and dark brown. Within the inaccurately predicted phenotypes, rs1800407 is confirmed to affect both inaccurately predicted groups and is the most significant SNP. Additionally, rs121918166, a missense variant in OCA2, is the second most significant SNP in true blue predicted as brown. Both SNPs were also the two most significant haplotypes with at least one allele being derived. Therefore, the next steps should include the addition of the functional haplotype and rs121918166 into the current IrisPlex model, and further testing of rs1448484 and rs61335644 on a molecular level. Consequently, the current IrisPlex model should also be reassessed on an independent test set using the 4-point categorical scale rather than the present 3-point scale.

La cryptographie visuelle consiste à partager entre plusieurs « Shadow Images » (SIs) un secret qui ne se révèlera à l'oeil de l'observateur qu'à leur superposition. Depuis les travaux de Naor et Shamir, ce procédé cryptographique a été étendu à Des schémas numériques variés, adaptés à diverses problématiques. En revanche, les travaux concernant son implémentation physique sont peu nombreux à ce jour. Cette thèse est consacrée à l'implémentation de la cryptographie visuelle sur des SIs imprimés en demi-tons en vue de l'authentification de documents. Le SI associé au document peut être imprimé sur support opaque ou transparent, les autres SIs étant imprimés sur films transparents. Nous avons résolu la difficulté de leur superposition par une méthode de Fourier permettant le recalage de la structure quasi-périodique d'un SI. La précision de cette méthode nous a permis de développer un système optique de superposition par projection. On verra que les phénomènes physiques responsables du rendu visuel de SIs superposés sont propices à une protection contre la copie illicite du SI associé à un document. La complexité de ces phénomènes et leur dépendance au type d'impression imposent une modélisation physique pour obtenir un rendu précis. Cette approche nous a conduit à considérer la problématique de la reproduction des couleurs et à développer des modèles spectraux adaptés à la superposition de supports imprimés non diffusants et/ou diffusants, en réflexion et en transmission. La précision de ces modèles prédictifs est tout à fait satisfaisante au regard de celle habituellement obtenue dans le domaine de la reproduction des couleurs. Cela nous a permis d'introduire une approche originale de la cryptographie visuelle par ajustement de couleur (color matching) : une même couleur, à une tolérance près basée sur la vision humaine, est générée par différents demi-tons imprimés sur les supports à superposer. La couleur du message peut ainsi constituer un secret partagé entre les SIs de la même façon que l'est le contenu du message. Chaque SI pris individuellement ne laisse fuir aucune information sur la couleur du message, qui ne sera révélée qu'à leur superposition. Cela peut permettre de prévenir une attaque par falsification du SI associé au document (cheating attack ). De plus, le rendu des couleurs étant très dépendant du système d'impression utilisé, une reproduction fidèle à partir d'un système d'impression différent est difficile. La difficulté peut être encore accrue par l'utilisation de caractéristiques d'impression non standard / In this thesis, we will focus on the physical implementation of visual cryptography, which consists in sharing a secret message between several unmeaning images, so-called shadow images, at least one of them being printed. By the principle of the method, no information leaks about the message until the images are properly stacked together. As the alignment of the shadow images hampers the deployment of the visual cryptography in practice, we develop a dedicated image registration method. In contrast with existing methods, ours is not intrusive. We make use of the particular shape of the elementary constituents of the shadow images, the shares, to extract in the Fourier domain the main parameters of the geometrical transformations occurring between the superposed images. We prove that this method allows subpixel accuracy in shadow images registration. We benefit from such ability by implementing visual cryptography in an image projection configuration : the digital shadow image is projected onto the printed one. In this way, the registration is performed automatically by using a digital camera (the resulting superposition being observable by the eye). For the purpose of authentication, one has to deal with specific attacks: the shadow image attached to a given document could be tampered with or copied. In order to prevent such attacks, we have increased the di_culty to reproduce the shadow image by considering color. This approach requires a complete management of colors. Thanks to recent advances in color reproduction, we are able to predict the reflectance and transmittance spectra of supports printed in color. In this thesis, we develop new spectral prediction models namely for piles of printed transparencies as well as for transparencies stacked onto papers, all printed in color. Thus, we are able to predict the color of each share in a shadow image to be printed and to achieve color matching i.e. we are able to reach a color by various combinations of superposed colors. Such a prediction allowed us to introduce a new approach in visual cryptography: color matching when revealing the secret message to be shared into two (or more) shadow images, in order to authenticate the shadow images provider. As the prediction models are sensitive to the calibration of the printing system (printer, inks, supports, halftoning and geometry measurement conditions), the use of special materials will increase the di_culty to generate visually acceptable fake pairs of shadow images