No description available.

No description available.

<p>To render outdoor scenes with lots of vegetation in real time is a big challenge. This problem has important applications in the areas of visualization and simulation. Some progress has been made the last years, but a previously unsolved difficulty has been to combine high rendering quality with abundant variation in scenes. </p><p>I present a method to mathematically generate and render vegetation in real time, with implementation in the scene graph Gizmo3D. The most important quality of the method is its ability to render scenes with many unique specimens with very low aliasing. </p><p>To obtain real time performance, a hierarchical level-of-detail scheme (LOD- scheme) is used which facilitates generation of vegetation in the desired level- of-detail on the fly. The LOD-scheme is texture-based and uses textures that are common for all specimens of a whole species. The most important contribution is that I combine this LOD-scheme with the use of semi- transparency, which makes it possible to obtain low aliasing. </p><p>Scenes with semi-transparency require correct rendering order. I solve this problem by introducing a new method for approximate depth sorting. An additional contribution is a variant of axis-aligned billboards, designated blob, which is used in the LOD-scheme. Furthermore, building blocks consisting of small branches are used to increase generation performance.</p>

High quality 3D rendering requires massive computing resources. In order to render animations within a reasonable amount of time, the rendering process is often distributed among a cluster of computers, typically called a rendering farm. However, most individuals and small studios do not have the resources to purchase or lease a rendering farm. In the late 1990?s, Java technology brought a hope that distributed applets could be utilized as an alternative to traditional network rendering models. Yet, this hope was never realized, nor was it fully implemented. Taking into account new developments in web application technology and the Sunflow renderer, this thesis will reexamine the possibility of distributed rendering applets. This thesis will suggest that distributed Java applets can effectively render projects across a collection of heterogeneous and geographically dispersed computers over the Internet. Moreover, this paper will present a prototype web application, called RenderWeb, that uses distributed applets to quickly render projects created in popular animation programs, such as Blender, 3D Studio MAX, and Softimage.

To render outdoor scenes with lots of vegetation in real time is a big challenge. This problem has important applications in the areas of visualization and simulation. Some progress has been made the last years, but a previously unsolved difficulty has been to combine high rendering quality with abundant variation in scenes. I present a method to mathematically generate and render vegetation in real time, with implementation in the scene graph Gizmo3D. The most important quality of the method is its ability to render scenes with many unique specimens with very low aliasing. To obtain real time performance, a hierarchical level-of-detail scheme (LOD- scheme) is used which facilitates generation of vegetation in the desired level- of-detail on the fly. The LOD-scheme is texture-based and uses textures that are common for all specimens of a whole species. The most important contribution is that I combine this LOD-scheme with the use of semi- transparency, which makes it possible to obtain low aliasing. Scenes with semi-transparency require correct rendering order. I solve this problem by introducing a new method for approximate depth sorting. An additional contribution is a variant of axis-aligned billboards, designated blob, which is used in the LOD-scheme. Furthermore, building blocks consisting of small branches are used to increase generation performance.

Embroidery is a traditional non-photorealistic art form in which threads of different colours stitched into a base material are used to create an image. This thesis presents techniques for automatically producing embroidery layouts from line drawings and for rendering those layouts in real time on potentially deformable 3D objects with hardware acceleration. Layout of stitches is based on automatic extraction of contours from line drawings followed by a set of stitch-placement procedures based on traditional embroidery techniques. Rendering first captures the lighting environment on the surface of the target object and renders it as an image in texture space. Stitches are rendered in this space using a lighting model suitable for threads at a resolution that avoids geometric and highlight aliasing. It is also possible to render stitches in layers to capture the 2.5D nature of embroidery. A filtered texture pyramid is constructed from the resulting texture and applied to the 3D object. Aliasing of fine stitch structure and highlights is avoided by this process. The result is a realistic embroidered image that properly responds to lighting.

In this thesis, an efficient font-rendering system for E-book applications was presented. The rendering of vector-type fonts mainly consists of two procedures: the rasterization of outline curves, and the filling of font interiors. Different from the conventional rendering systems which approximate the curve using many tiny straight lines, our system applies the fast adaptive forward-difference algorithm to render the curve. In addition, for the filling of fonts, this thesis adopts the local-minimum table and the vertex table to substitute the conventional global edge table in order to reduce the memory requirement. Finally, in order to improve the rendering quality of the fonts, our system has applied the super-sampling anti-aliasing technique based on the FLIPQUAD sample patterns used in three-dimensional graphic systems in order to reduce both the required buffer size and the filling operation complexity by 25% for the similar rendering quality. Our rendering system has been implemented in an ARM-based platform. The FreeType fonts parser has been used, and transformed into the application programming interface (API) commands of a two-dimensional vector graphics standard called OpenVG. This standard APIs have been built on the QT window and Embedded Linux operational systems. The rendering system proposed by this thesis represents a total open-source solution, and can be applied to many embedded applications.

This thesis presents an expressive (non-photorealistic) rendering approach created by combining volumetric rendering techniques with the Line Integral Convolution (LIC) in 3D space. Although some techniques that combine volume rendering with the LIC have been introduced in computer graphics, they are mainly used for the scientific visualization fields, such as the visualization of 3D fluid fields. Unlike earlier research, we will focus on artistic representation, which is significantly different than scientific visualization research. We will implement a brush-stroke effect on the implicit surfaces by using the LIC. The implicit surfaces are described as volume datasets that are created by the voxelization of triangular meshes. To acquire smearing effects on the surface we convolve along the vector fields with the densities of the voxels of the datasets. These vector fields are defined by users as texture maps. The final images are rendered with volume ray casting, integrating colors and densities of voxels with Perlin noise along vector fields. The Perlin noise provides randomness and allows us to generate scratches. Smearing effects on the surface of an object create the illusion of 3D brush-strokes as if a painter had created brush strokes on a canvas. The rendering system is implemented using standard C and C++ programming languages. 3D models are then created using Alias MayaTM and TopmodTM.

This thesis presents a general method and guidelines for compositing digital characters into traditional artwork by matching a character to the perspective, lighting, style, and complexity of the particular work of art. The primary goal of this integration is to make the resulting image believable, but not necessarily to create an exact match. As a result, the approach used here is not limited to a single rendering style or medium, but can be used to create a very close match for almost any artistic image. To develop and test this method and set of guidelines I created composites using a variety of styles and mediums.

Dažnai kompiuterinės grafikos taikymuose labiausiai visos scenos atvaizdavimo greitį įtakoja kraštovaizdžių atvaizdavimas. Darbe nagrinėtas vienas iš dviejų greičiausių šiuolaikinių kintamo detalumo kraštovaizdžių atvaizdavimo algoritmų – geomipmap. Šis algoritmas pasirinktas dėl platesnio pritaikomumo, didesnio algoritmo lankstumo ir šiai dienai dažnai didesnio greičio lyginant su jo varžovu geoclipmap. Algoritmo idėja – panaikinti kas antrą viršūnę iš vaizduojamųjų sąrašo, pereinant prie žemesnio detalumo lygio, taip sudarant galimybę laikyti visą geometriją keliuose nekintančiuose viršūnių ir indeksų buferiuose. Darbe pasiūlyta algoritmo optimizacija išnaudoja iš aukščių žemėlapio generuojamų kraštovaizdžių apatinę ribą. Pagrindinė idėja – nevaizduoti visų šioje riboje esančių trikampių, vėliau visus atsiradusius tarpus uždengti naudojant vieną plokštumą, ištemptą per visą kraštovaizdį. Nors optimizacijos efektyvumas priklauso nuo kraštovaizdžio formos ir negalima vienareikšmiškai įvertinti atvaizdavimo pagreitėjimo bendru atveju, nagrinėtose scenose užfiksuotas didelis našumo padidėjimas, aiškiai vertingas, net jei pasireikštų tik labai retais atvejais. Darbe taip pat parodoma, kad našumo padidėjimas iš tiesų galimas daugelyje scenų. / Often the bottleneck in the real-time 3D rendering applications is the rendering of a terrain. Geomipmap algorithm is analyzed here as it is one of the best terrain rendering algorithms known. It is much more versatile and flexible in comparative to geoclipmap and still often outperforms its main rival. The idea is to eliminate every other vertex in order to get a lower level of detail, thus having all the geometry in several immutable vertex and index buffers. An optimization is proposed that utilizes the lower bound of terrains generated from a heightmap. The idea is not to render the triangles that are exactly on the lower bound of a terrain and to cover all the eliminated triangles with a single plane. Even though the effectiveness of the optimization is highly dependent on the exact form of the terrain, the high performance gain observed in the analyzed scenes is valuable even if it would only occur in very rare cases. It is shown that the performance gain can actually be observed in quite a variety of scenes.