Ce mémoire de maîtrise porte sur l'interrelation existante entre les images et les sons qui composent des productions culturelles électro-numériques télévisuelles. Cette recherche a pour but d'observer comment le traitement informatique influence le rendu d'une publicité télévisée. La méthodologie utilisée, qui repose sur une grille d'analyse de caractéristiques techniques et iconiques, permet l'observation de l'ensemble de la production et de son évolution temporelle. Le terme iconique inclut dans notre mémoire tout autant l'image que le son. S'ensuit une analyse de la composition de la publicité et de ses spécificités sémiologiques. Cela permet en dernier lieu de classifier la production selon son caractère d'homogénéité. Ces observations permettent une appréciation de chaque publicité et la critique des interrelations entre le niveau visuel et le niveau sonore présentes au sein de ces productions. Le corpus est constitué de quatre productions publicitaires créées par trois agences. Les publicités qui ont été privilégiées démontrent un degré manifeste de traitement électro-numérique. Cela implique que l'utilisation de logiciels informatiques y est perceptible. Ces productions contiennent au moins un élément visuel ou sonore qui sera d'origine électro-numérique. Elles ne présentent pas toutes le même degré de traitement ce qui permet l'observation d'un éventail de possibilités concernant l'utilisation de l'informatique en création. Préalablement à l'analyse du corpus, les sujets du monde publicitaire, des techniques de traitement de l'image et du son et de l'interdisciplinarité sont discutés. Ces thèmes sont importants dans l'analyse de notre corpus. L'hypothèse de cette recherche repose sur l'idée qu'un son, à un instant donné, doit faire preuve d'un traitement électro-numérique compatible au traitement de l'image simultanée. Si c'est le cas, le rendu global sera homogène et il n'y aura pas obstruction du message. Cependant, l'analyse du corpus démontre que selon la nature de l'objet de la publicité, diverses possibilités sont envisageables. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : Publicité, Électronique, Numérique, Électro-numérique, Abstraction, Langages homogènes, Production culturelle.

Volume visualization is one important part of scientific visualization. It has developed basing on absorbing the relative knowledge of computer delineation, computer visualization and computer image disposal. The knowledge of this branch is of much importance in computer application. Since it deals with contents with deeper meaning and it is more theoretic, having more arithmetic means, it generally stands for the level of computer application. The study and application of volume visualization is like a raging fire. My country started comparatively later in this field. This thesis gives systematic representation and discuss in the field of tomography image 3D reconstruction. It mainly discusses after rotation, translation, filtering, interpolation and sharpening a series of 2D CT scanning images, get the boundary data of different object to form 3D volume data and actualize the 3D reconstruction of the object, and at the same time implement the function of display, translation, rotation, scaling and projection the volume data. Basing on the implementation of these functions according to software programming, this thesis gives a sum up to each algorithm of 3D volume visualization processing. The method to actualize the 3D reconstruction of the tomography image is mainly about the image processing, image transformation, the way to actualize 3D reconstruction and image compression, etc. In image processing, it talks about getting the anchor points in the tomography image, the geometrie transformation of the image, getting the boundary of the target, cross section display and the smoothing and sharpening of the image. In the transformation of the image, this thesis deals with the algorithm and implementation principle of the geometric transformation (transition, rotation, and scaling) of the 2D image, the three-dimensionalization of the planar data, construction of the stereo mode, geometric transformation of the 3D graph, curve-fitting, the processing of hidden line and hidden surface, color processing. It also introduces the thoughts of using OpenGL to develop and actualize tomography image 3D reconstruction system, including using OpenGL to transform the coordinate, solid model building, to actualize 3D rotation and projection. Recently, the methods of applying chemotherapy to deal with cancer in hospitals of our country are different. Hospital with great fund takes import software to design while most of the hospitals take domestic software. These kinds of software are designed by DAHENG Company in BeiJing, WEIDA Company in ShenZhen. The programs in the software these two hospitals' doctors use to treat are images fielding in the plane not making radiation design under the mode of omnibearing cubic display. Under this circumstance the judgment of the key part can not remain precise, and this part is the most important part that the doctors need. The above problem is the aim of this project. This thesis mainly deals with the subject that after calibrating and sharpening the series of 2D CT images, extract the boundary data of different bodies to form a 3D volume data and actualize 3D reconstruction and at the same time actualize the function of display, translation, rotation, scaling and projection. Mostly basing on the application of medical area, this thesis aims at making further research on computer graphies, computer vision and computer image processing through the study and application of volume visualization in this field. By the study and development of the volume visualization technology in this project, it can provide simulation and display functions to the observer before the operation and the radiotherapy as well as providing the chance to simulate the real teaching and practicing link to the medical school in the teaching process, and increase the clinic level and teaching level of medical area. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : volume visualization, volume data, tomography image, 3D reconstruction, anchor point, boundary data, periphery boundary, OpenGL.