This work presents a method to extract polygonal surfaces from volumetric models created by artists, proposing a way of using voxel modeling tools to build B-Rep models.

   The volumetric data created by voxel editors or obtained from processes that acquire 3D geometry from volumes may contain topological features does not describe solid structures, because they do not have a frontier that separates the interior and the exterior of the volume in a clear and well defined way. The main objective of this work is to solve the problem of extracting triangle meshes from volumes that contains these topological features.

   In order to extract surfaces successfully, a methodology was conceived to resample any volumetric model, in a way that it is possible to reconstruct a tridimensional manifold that can be polygonized without generating surface with gaps or topological problems.

   The mesh generated by this technique is smooth and its triangles have good properties, satisfying some of the main criteria used to measure the quality of meshes, such as aspect ratio, smoothness and skewness.

   Despite having triangles with good properties, the models extracted by this method may present more faceted features than what was expected. In order to avoid this problem, it is necessary to apply a method to obtain a surface with a smoother geometry that preserves crucial details of the original model. The mesh smooth method proposed and used at this work guarantees that the produced surface is still a 2-manifold by not introducing degenerations at the topological structure of the generated object.

   This work presents a study on the reconstruction of polygonal surfaces from volumetric models, striving to devise new way of building graphical objects mixing voxel manipulation techniques with vertex based modeling, available in 3D image editing applications. While the use of boundary representation is recommended for constructing complex and large meshes, for some applications as, for example, videogames, the manipulation of voxels is more intuitive in the process of adding details to models, providing a more accurate control to the user than function based textures. The method developed in this work conducts the process of exporting volumetric pixels into sets of faces, edges and vertices, detects the surface normals, smooths the colors and the geometry of the surfaces and also reduces the polygon count of the generated mesh. This technique was tested with the help of the program Voxel Section Editor III, written in Delphi programming language, which has a vast amount of free graphical models, being constantly created using the file format supported by it.

   Geometric modeling has recently leveraged the power of Generative Design. Generative Design can be seen as a framework in which models can be generated by systematically exploring a space of shapes generated by recombination of parametric shape descriptors. In this work, we explore the use of Generative Design for modeling 3D clouds. Clouds are usually modeled in Computer Graphics as voxelized models by using a combination of implicit and procedural techniques. Hence, they are described by a large number of parameters. Although these parameters usually include parameters that define a combination of scalar fields, noise functions and affine transforms, the controlled use of such parameters is rather complex. How to tune them up to obtain a plausible result is not obvious. We propose a method based on generative design combined with Machine Learning to produce families of cloud shapes automatically. Our generative design method is based on an evolutionary approach that generates instances of plausible 3D cloud shapes by optimizing a fitness function that measures the likelihood of a shape be a cloud. As the manual design of a fitness function is also quite complex, we propose using a Convolutional Neural Network to learn the fitness of arbitrary 2D views of the generated clouds. We perform several experiments that confirm the viability of the proposed method compared to manually modeled clouds.

   This work presents a method to extract polygonal surfaces from volumetric models created by artists, proposing a way of using voxel modeling tools to build B-Rep models.

   The volumetric data created by voxel editors usually contain topological features that do not describe solid structures. Hence, the main objective of this work is to solve the problem of extracting triangle meshes from volumes that contain these topological features.

   In order to extract surfaces successfully, a methodology was conceived to resample any volumetric model, in a way that it is possible to reconstruct a tridimensional manifold that can be polygonized without generating a surface with gaps or topological problems.

   The meshes generated by this technique have good properties, satisfying some of the main criteria used to measure the quality of meshes, such as aspect ratio, smoothness and skewness.

   This work proposes a new method for the polygonization and texture map extraction from volumetric objects based on a polyhedral surface extraction and surface fairing approach. The surface faring approach is based on a mesh signal processing technique that uses an approximation of an adaptive low-pass filtering based on a function that combines the discrete laplacian operator and an adaptation of the Lanczos kernel. Differently from other works that propose surface fairing approaches we deal mainly with voxelized models created by artists for games which led us to design a filtering approach that smoothes strongly in high frequency regions to remove the jagging effects of the sampling and less intensively in the low frequency regions to preserve the natural behavior of the model. The overall aim of this work is to propose a simple methodology in which it is efficient and easy to use voxel manipulation techniques to produce a final boundary representation of the model together with the appearance function codified in a texture atlas. The proposed method produced very promising results and enabled us to extract smooth texturized models that preserve the features of the original volumetric object at interactive times.

   Volumetric rendering is a way to represent 3D models on devices with a low use of 3D resources. It allows the use of high poly models, quick collision detection and rasterization operations, making them interesting for the games that are being created for the current generation of mobiles and pocket pcs that cannot handle OpenGL operations with a fast frame rate. However, the lack of knowledge of the shape of the object has proven to be a problem to preprocess its lighting. This short paper presents an heuristic that finds an approximation of the direction of normal vector of a voxel without the previous knowledge of the mesh, vertexes, edges and surfaces of the volumetric model where it belongs.