Geoff Wyvill

Data Structure for Soft Objects

We introduce the concept ofsoft objects whose shape changes in response to their surroundings. Established geometric modelling techniques exist to handle most engineering components, including ‘free form’ shapes such as car bodies and telephones. More recently, there has been a lot of interest in modelling natural pheomena such as smoke, clouds, mountains and coastlines where the shapes are described stochastically, or as fractals. None of these techniques lends itself to the description ofsoft objects. This class of objects includes fabrics, cushions, living forms, mud and water. In this paper, we describe a method of modelling such objects and discuss its uses in animation. Our method is to represent asoft object, or collection of objects, as a surface of constant value in a scalar field over three dimensions. The main technical problem is to avoid calculating the field value at too many points. We do this with a combination of data structures at some cost in internal memory usage.

Analysis of an algorithm for fast ray tracing using uniform space subdivision

Ray tracing is becoming popular as the best method of rendering high quality images from three dimensional models. Unfortunately, the computational cost is high. Recently, a number of authors have reported on ways to speed up this process by means of space subdivision which is used to minimize the number of intersection calculations. We describe such an algorithm together with an analysis of the factors which affect its performance. The critical operation of skipping an empty space subdivision can be done very quickly, using only integer addition and comparison. A theoretical analysis of the algorithm is developed. It shows how the space and time requirements vary with the number of objects in the scene.

SIFT and SURF Performance Evaluation against Various Image Deformations on Benchmark Dataset

Scene classification in indoor and outdoor environments is a fundamental problem to the vision and robotics community. Scene classification benefits from image features which are invariant to image transformations such as rotation, illumination, scale, viewpoint, noise etc. Selecting suitable features that exhibit such invariances plays a key part in classification performance. This paper summarizes the performance of two robust feature detection algorithms namely Scale Invariant Feature Transform (SIFT) and Speeded up Robust Features (SURF) on several classification datasets. In this paper, we have proposed three shorter SIFT descriptors. Results show that the proposed 64D and 96D SIFT descriptors perform as well as traditional 128D SIFT descriptors for image matching at a significantly reduced computational cost. SURF has also been observed to give good classification results on different datasets.

Space Division for Ray Tracing in CSG

A system of Constructive Solid Geometry (CSG) enables an engineering designer to compose threedimensional shapes by combining simpler ones. Definitions of such objects are represented by tree structures or directed acyclic graphs. Most existing systems convert this representation to a more conventional boundary representation of the solids in order to render pictures from the model. More recently, a number of systems have been described that render the pictures directly from the CSG structure. We describe such a system. We render a scene by ray tracing from a directed acyclic graph. This process is made efficient for large models by using an adaptive method of space division to reduce the number of intersection calculations needed.

Field functions for implicit surfaces

The use of 3D computer generated models is a rapidly growing part of the animation industry. But the established modelling techniques, using polygons or parametric patches, are not the best to define characters which can change their shape as they move. A newer method, using iso-surfaces in a scalar field, enables us to create models that can make the dynamic shape changes seen in hand animation. We call such modelsSoft Objects.From the users point of view, a soft object is built from primitive key objects that blend to form a compound shape. In this paper, we examine some of the problems of choosing suitable keys and introduce some new field functions that increase the range of shapes available as keys.

A functional model for constructive solid geometry

A system of constructive solid geometry (CSG) enables an engineering desiger to compose three dimensional (3D) shapes by combining simpler ones. Most existing systems, however, actually represent solid shapes as boundaries of surface patches. At the Kunii Laboratory, University of Tokyo, we have produced an experimental system in which solids are modelled functionally by procedures which describe their properties. These “primitive objects” are combined with the aid of a new “octree” structure. Careful study of the data structures in this system reveals some interesting aspects of program efficiency.

Swirling-sweepers: constant volume modeling

Swirling-sweepers is a new method for modeling shapes while preserving volume. The artist describes a deformation by dragging a point along a path. The method is independent of the geometric representation of the shape. It preserves volume and avoids self-intersections, both local and global. It is capable of unlimited stretching and the deformation can be constrained to affect only apart of the model. We argue that all of these properties are necessary for interactive modeling if the user is to have the impression that he or she is shaping a real material. Our method is the first to implement all five.

Ray-tracing soft objects

Soft objects, also known as metaballs or implicit surfaces, are deformable free-form shapes represented as a surface of constant value in a scalar field.

Swirling-sweepers: constant-volume modeling

Swirling-sweepers is a new method for modeling shapes while preserving volume. The artist describes a deformation by dragging a point along a path. The method is independent of the geometric representation of the shape. It preserves volume and avoids self-intersections, both local and global. It is capable of unlimited stretching and the deformation can be constrained to affect only apart of the model. We argue that all of these properties are necessary for interactive modeling if the user is to have the impression that he or she is shaping a real material. Our method is the first to implement all five.

Solid Texturing of Soft Objects

Since the shape of a soft object changes in response to its surroundings, it is difficult to give a single position in space as the location of the object. Indeed objects can and do break dynamically into subobjects. This means that you cannot map a solid texture onto such an object simply by using a function of the space coordinates. We have taken a different approach. Our soft objects are modeled as the volume enclosed by an isosurf ace of a field calculated from a set of key points. For each key point, we describe an abstract texture space. Any point on the surface of an object has, associated with each key point, a position in this space and a field contribution. A vector sum of these positions, weighted by the field contributions, is used to select a surface specification from the texture space. Textures mapped with this process retain their consistency during distortion and metamorphoses of objects, permitting a great variety of animation effects.