New Trends in Computer Graphics

Locating sampling points for cubic splines is investigated in this paper. The direct application of the proposal is for designing motion of objects for computer animation. The method consists of two parts: (i) obtaining the relative position of the sampling points, and (ii) mapping the sampling points on the cubic spline interpolant based on the relative position sequence obtained in (i).The obtained results may be used for design of motion for phenomena described parametrically.

One of the biggest problem in computer animation is the enormous amount of specification necessary to produce an animation sequence. Our animation language, SOLAR, is designed to enable us to specify animation easily and efficiently. SOLAR achieves this through the use of object-oriented paradigm and abstractions. Supports for high-level abstractions and adaptive motion are provided through its class-inheritance and message-passing mechanisms. Five kinds of abstractions are supported, i.e. structural, motion, functional, character and world modeling. Together, they provide good support for programming animation. All interactions between graphical objects are achieved through its messagepassing mechanism. This paper describes the main features of SOLAR and its implementation.

The present paper emphasized a total approach to 3-D computer animation, based on AVENUE. An integrated animation system needs to support all kinds of animation, and AVENUE has this potential, since it supports smooth object modeling, motion specification, and rendering. The authors describe the system concept[1], a modeling method using a geometry language, and a motion specification method using a 3-D interactive parametric keyframe and rule based motion analysis. They also explain a rendering technique that uses sub-division technique for CSG(Constructive Solid Geometry) with deformation.

Quaternions, although not well known, provide a solid base to describe orientation of an object or a vector. They are efficient and well suited to solve rotation and orientation problems in computer graphics and animation. This paper describes a new method for splining quaternions so that they can be used with keyframe animation. We also show that quaternions, although up to now solely used for animation purposes, can be used succesfully in the field of modelling and rendering and we prove that we can construct a significantly faster rendering algorithm with the use of quaternions.

In this paper we present an analysis of animation control processes. The design of an animation system is strongly coupled with the considered application. So these systems are specialized, some of them in audiovisual production, in C.A.D. or in mechanical computation. We are working on the design of a general animation system with the capability to generate motion resulting from mechanical laws in an audiovisual environment. The heart of this system is a structured graph used to store a hierarchical description of the objects and the mechanical joints linking them together to build a multibody mechanical system. Motion control can also be specified by key-framing techniques or explicit trajectories for the objects which are not submitted to mechanical laws. The dynamical formalism takes into account holonomic and nonholonomic constraints using the principle of virtual works associated with LAGRANGE’s multipliers. Symbolical equations of motion are automatically built by the system and solved for each time step (frame) to give object locations and orientations.

Most established methods using three dimensional (3-D) computer graphics for human facial animation fall short of achieving truly realistic images, due to the lack of sufficient surface detail for the eyes, mouth and skin/hair textures. A new method is presented here, combining 3-D computer graphics with texture mapping techniques to synthesise and animate facial images which are convincing enough to be comparable with real video images. Animation is achieved by manipulation of a 3-D model of the subject’s face in conjunction with a set of pre-stored digitised image samples.

Creation of new synthetic actors is a tedious and painful task. The situation may be improved by introducing tools for the creation. Three approaches are discussed in this paper: modification and edition of an existing synthetic actor using local transformations; generation of new synthetic actors obtained by interpolation between two existing actors; creation of a synthetic actor by composition of different parts. Animation of synthetic actors is also discussed.

The goal of the project “Occursus Cum Novo” was to generate a complex photo-realistic animation of nontrivial length in reasonable time at reasonable costs. Photographic realism comprises complex geometric models as well as simulation of several optical effects. This paper starts with an introduction to the “Occursus Cum Novo” modeling environment. Following the toolkit approach, it offers a set of tools for textual, graphical interactive, and simulative modeling, embedded in the UNIX programming environment. The second part is devoted to rendering. Photo-realistic pictures generated by raytracing are still those of highest quality. However, due to the tremendous time of computation, raytraced animations are rather rare. An organization scheme for rendering on a network of work stations is described which enabled us to generate a 5-minutes raytraced animation within 2 months without affecting any of the regular users of the work stations. The results of the project are of general interest since they show a way for efficient high quality photo-realistic animation synthesis for the future.

This paper attempts to generalize and to unify the shading models described by Gouraud [1] and Phong [2]. The major effort in this direction is to produce a model that exhibits the following features: 1.It is incremental in an arbitrary scan direction across a two dimensional polygon.2.The increments in the x and y directions are constant across the elementary polygon to be shaded.3.It maintains continuity of the intensity I or of the normals N when traversing the edges that bound the elementary polygons.4.In the case of the shading process following the projection it considers the contribution of the z coordinate.5.The shading is invariant with respect to rotation.The first three features are satisfied by the models proposed by Gouraud and Phong whereas the last two are not. It is the purpose of this paper to introduce a general model that is invariant with respect to rotation, can be applied either before or after projection and is insensitive to the presence of concave vertices.

In computer graphics curved surfaces are commonly represented by a mesh of polygons. When these objects are rendered, the ‘visibility’ of the polygonal definition is diminished by using an interpolative shading algorithm. The two algorithms used for this are Phong shading (Phong 1975) and Gouraud shading (Gouraud 1971). Phong shading is capable of producing more realistic specular highlights, but is more expensive than Gouraud shading.In general for an object, the number of specularly highlighted polygons is small compared to the number of non-highlighted polygons, especially for a single light source. We describe a reliable and consistent method which allows objects to be Gouraud shaded, with Phong shading used only to add the specular highlights. Thus, images of the quality associated with Phong shading can be produced with little more computation than that needed for an equivalent Gouraud shaded image.

The algorithm for synthetic image generation with the lens effect consists of two consecutive processors: the hidden-surface processor and the focus processor (Potmesil and Chakravarty 1981). This algorithm was revised by Chen (1987) using simple light particle theory instead of the wave theory to avoid the complicated calculation and the huge memory consumption. However, from the experimental observation, the algorithm based on hidden-surface processor and focus processor can only serve as an approximation formula for the slightly defocused scenes. The purpose of this paper is to present a new algorithm for the highly defocused scenes. An image generated using this new algorithm was compared to the real photograph captured by a camera, and no noticeable difference of the defocused effect between the image and the photograph can be detected.

An artistic approach for modeling fuzzy objects called Particals is presented. The method unifies the modeling and rendering process. Particals treat fuzzy objects as a cluster of miniscule particles. The distribution of particles is defined by a stochastic function. Particles are generated during the rendering process by referencing the stochastic function. Ray tracing is used for rendering so that intersection, reflection, refraction, and shadow calculations can be incorporated. The algorithm is efficient and it produces impressionistic images with very intricate textures. However, it suffers from the stobing effect in animation sequences because the method does not maintain frame to frame coherence. The particles appears to be equal in size because the size is fixed to the size of the pixel on the image screen.

The ray-tracing algorithm was up to now the origin of the most beautiful synthesised images. Yet, it was penalized by the great number of intersection computations required, due to the non-exploitation of the spatial coherency.In order to decrease the cost of image production, architectures were proposed, based on the ray-tracing algorithm parallelism.We propose a volumic architecture model traced over reality, using the parallelism due to the scene coherency.This architecture is based on the object decomposition into voxels and on the incremental integer logic. This allows the suppression of the intersection computations at the rendering step.During the object decomposition into voxels, the texture is included into the material features. Thus, the objects are not created plain but textured just as if they were sculptured into the material.The object decomposition into voxels allows the generation of composite scenes where the objects modelled by different methods will be rendered by the same technique.

Associative processor arrays are described in relation to a VLSI processor array currently being designed at Bristol University. Its application to a number of image generation tasks is considered. Some performance estimates are given for a processor-per-pixel system incorporating this chip

We present a systematic and efficient strategy for mapping an adaptively/regularly subdivided object space (a set of subspaces) into the nodes of the hypercube. The property of this mapping is that the distance between the neighbouring subspaces on the hypercube is proportional to the difference between the sizes of these subspaces. Especially, if neighbouring subspaces are of equal size, these subspaces are allocated to the neighbouring processors. As a result, we can realize a communication-effective implementation of parallel ray-tracing on the hypercube multiprocessor system. The mapping is derived from the byproduct of octree encoding of an object space.

A parallel space tracing algorithm is presented. It subdivides the scene into regions. These latter are distributed among the processors of a 2D array architecture which is mapped onto an iPSC hypercube machine designed by Intel company. Each processor subdivides its own region into cells to accelerate the ray tracing algorithm. Processors communicate by means of messages. The pyramidal shape of the regions allows to delete the primary ray messages. An efficient termination algorithm is described. A method of performing a roughly uniform load distribution is proposed.

A constructive solid geometry for simulation models is composed of fundamental primitives in a ray tracing algorithm. Only a ray of light from two components of specular reflection and refraction is simply traced backwards according to the value of depth level and threshold of energy attenuation. Ray traced images of a main transparent object enclosed with another transparent object are displayed on a color CRT. Complex color rendering view of double transparent objects is discussed from a viewpoint of the color shift and visual transparent quality of computer-generated images.

Aliasing is a decisive problem in realistic image producing. Since ray tracing is a rather slow algorithm of visualization, antialiasing an image by systematically oversampling its pixels is quite costly. We suggest a local adaptive oversampling algorithm for antialiasing ray tracing. We use space coherence to determine which pixels on the screen need be oversampled. The C.S.G. model is used to define the scene, and also to limit it, for each pixel, to objects of concern only. To minimize memory space needs, we work on a window over the screen.CR Categories and Subject Descriptors: 1.3.3 [Computer Graphics]: Picture/Image Generation; 1.3.5 [Computer Graphics]: Computational Geometry and Object Modeling; 1.3.7 [Computer Graphics]: Three-Dimensional Graphics and Realism.

Image warping refers to the 2D resampling of a source image onto a target image. Despite the variety of techniques proposed, a large class of image warping problems remains inadequately solved: mapping between two images which are delimited by arbitrary, closed, planar curves. Such problems are typified by the requirement to derive a spatial transformation given only boundary correspondence information.This paper describes a novel algorithm to perform image warping among arbitrary planar shapes whose boundary correspondences are known. A generalized polar coordinate parameterization is introduced to facilitate an efficient mapping procedure. Images are treated as collections of interior layers, extracted via a thinning process. Mapping these layers between the source and target images generates the 2D resampling grid that defines the warping. This mapping is shown to be decomposable into three ID transformations, thereby highlighting the primary benefit of the new parameter space. Applications include fast nonrectangular convolution, elastic matching, quantification of shape deformation, and visual effects.

A new method for filling a color table is presented that produces pictures of similar quality as existing methods, but requires less memory and execution time. All colors of an image are inserted in an octree, and this octree is reduced from the leaves to the root in such a way that every pixel has a well defined maximum error. The algorithm is described in PASCAL notation.

In this paper, we present a polygon handling technique for two-dimensional applications, based on a boundary representation, namely a winged-edge data structure. First, properties of boundary models modified to represent planar elements are analyzed, by extending the Euler formula to the case of faces lying in a plane.The concept of history is used as a connection between the produced output polygons and the input polygons, so providing a powerful retrieval tool. Finally, a simplified data structure is presented, as a friendly interface between the user and the internal polygon representation.

If we use a red drill to make a hole in a blue block, the inside of the hole will still be blue. To simulate this process in a CSG system, we subtract a red cylinder from the blue block. The hole’s surface belongs to the cylinder but it takes its colour from the block.We present algorithms for elucidating the correct colours from CSG models and describe an application in design for wood turning.

Building larger objects out of smaller rectangular parallelepipeds may be called a stacking operation. A modeling technique is developed for stacking a given number of parallelepipeds in a predefined arrangement. The technique uses the method of spatial occupancy and takes advantage of the specific properties of the stacking order for obtaining a solid model with hidden lines/surfaces removed. The algorithm uses efficient methods for handling huge databases and large computation time needed for solid modeling on microprocessors.

Set operation evaluation is an important component of Geometric Modeling System (GMS). Input data can be a CSG (Constructive Solid Geometry) representation with solid primitives approximated by polyhedra. This representation is a powerful tool for solid object description. Output data will be the unique resulting polyhedral object which provides an efficient data structure in display field. With no use of spatial coherency, computational complexity of set operation is quadratic. In this paper, we are introducing a new space subdivision scheme called Boolean Octree which performs set operations on polyhedra in an efficient way. This structure aims at limiting set operation evaluation in a ‘minimal space of calculation’ where primitive boundaries intersect each other and where resulting evaluation participates in the construction of the final resulting object. Boolean Octree computes set operations in a local level providing a linear complexity for geometric calculations. During space subdivision, Boolean Octree has a global view on local CSG tree (projection of the CSG tree in local space) taking into account simplifications of the boolean expression, avoiding evaluation and subdivision for the object parts out of the ‘minimal space of calculation’.

The uniform grid data structure is a flat (non-hierarchical) grid whose resolution adapts to the data. An exhaustive analysis of the uniform grid data structure for determining intersections in a set of many small line segments is presented. Databases from cartography, VLSI, and graphics with up to 115,973 edges are used. For each data set the intersection time, the ratio of edge pairs tested to pairs found to intersect, and size of intermediate data structures was measured as a function of grid resolution. The execution time was relatively insensitive to the grid size over a range of up to a factor of 10. 115,973 edges were processed to find 135,050 intersections in 683 seconds on a Sun 3/50 workstation. This data structure is also ideally suited for implementation on a parallel machine. When executing on a 16 processor Sequent Balance 21000, total times averaged ten times faster than when using only one processor. Finding all 81,373 intersections in a 62,045 edge database took only 28 seconds elapsed time. This research shows that more complicated, hierarchical data structures, such as quadtrees, are not necessary for this problem.

A variable-resolution boundary model of three dimensional objects provides a representation of the surfaces enclosing it at successively finer levels of specification. A hierarchical graph structure, called a structured edge-face graph (SEFG) is described, which encodes the boundary of a solid object at variable resolution is described. The structured edge-face graph is based on a face-oriented relational description of the object boundary. The concept of adjacency relation between pairs of primitive boundary entities at fixed accuracy is introduced, and adjacency finding algorithms which operate on a structured edge-face graph object representation are presented.

No one has yet been able to triangulate a simple polygon of n vertices in O (n) time. The fastest algorithm to date, due to Tarjan and van Wyk, runs in 0 (n loglogn) time. On the other hand several classes of simple polygons do admit linear-time triangulation. Some examples of such famous classes are: star-shaped, monotone, spiral, edge visible, and weakly externally visible polygons. In this paper the notion of geodesic paths is used to characterize all the classes of polygons for which linear time triangulation algorithms are known. First we introduce a new class of polygons, termed palm polygons, which subsumes many known classes of polygons for which linear time triangulation algorithms are known, and present an algorithm for triangulating palm polygons in 0(n) time. Then a class of polygons termed crab polygons is defined and shown to contain all classes of existing polygons for which linear time triangulation algorithms are known. As a by product of this characterization we obtain a new very simple linear time algorithm for triangulating star-shaped polygons.

Conventional scan-conversion algorithms were developed independently of filling algorithms. They cause many problems, when used for filling purposes. However, today’s raster printers and plotters require extended use of filling, especially for the generation of typographic characters and graphic line art. A new scan-conversion algorithm, called vertical scan-conversion has been specifically designed to meet the requirements of parity scan line fill algorithms. Vertical scan-conversion ensures the selection of exactly one pixel per intersecting scan line between a local minimum and a local maximum of the shape outline. Pairs of selected pixels define horizontal spans. All horizontal spans contain the full set of pixels interior to the original shape. Vertical scan-conversion greatly simplifies traditional edge-tracking filling algorithms, such as ordered edge fill, flag fill and descriptive contour fill, removing the need for testing and processing special cases.

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 models, Soft Objects. From the user’s 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.

Two cellular arrays are presented in this paper that are able to compute bicubical Bspline polynomial coefficients. The first array can be used to compute in parallel all the coefficients for a given surface and as well as providing a speedup factor of 256 compared to the single processor computation. The second array allows to partition the computation of the coefficients so that a smaller size of the array is required. This allows the user to reach a reasonable tradeoff between the speed needs and the VLSI implementation requirements.

The tensor product Bézier patch is one of the most widely used models for the representation of surfaces in CAGD. The shape of the objects to be designed is often complex so a piecewise representation is needed in most cases. If the resulting piecewise surface is to present a smooth aspect of its shape, then the geometric continuity between adjacent surface patches is essential.Much research has been devoted to this problem, and diverse solutions have already been published. Nevertheless, various problems remain, especially the smooth connection between a non-four-number of patches meeting at a common corner. In this paper, we first study the general behaviour of the G1 continuity constraints around an N-patch corner, which leads to useful results concerning the propagation of these constraints according to the parity of N. Then we present the conditions which allow a local determination of the surface patches, and analyze the remaining degrees of freedom which can be used to modify locally the surface shapes. These results are very useful for the design of a piecewise representation of smooth complex surfaces using Bézier patches where various configurations must be used for the connection at a corner of a different number of patches.

A new principle of generating three-dimensional shapes from two-dimensional images based on generalized symmetry is presented. Generalized symmetry is an extended concept of symmetry and represents the symmetrical characteristics of an object in terms of its curvilinear symmetry axis. This paper first develops the definition and the constraint of the generalized symmetry, and then describes an algorithm which generates the three-dimensional shape of an object from a line-drawing. This assumes that the line-drawing is an orthographic projection of an object which exhibits generalized symmetry. Several experiments by computer simulation verify that the algorithm can generate three-dimensional shapes from line-drawings.

This paper presents an interactive graphics system called L.E.G.O. The purpose of L.E.G.O. is to model two- and three-dimensional objects using Euclidean geometry constructions. Constructions are described by programs in the L.E.G.O. language, based on LISP. These programs can be entered in textual form or developed using a graphical interface in a multiple-window environment. Both interface styles can also be used concurrently. Applications of L.E.G.O. include computerassisted instruction of geometry, geometric modeling, and kinematic analysis. The use of imperative constructions and the powerful interface based on the idea of graphical programming are the most distinctive features of the system.

The ability to exchange standard data among various computer-aided design (CAD) and computer-aided manufacturing (CAM) tools used in discrete parts manufacturing provides an important key with regard to the full implementation of computer-integrated manufacturing (CIM). This paper will discuss two coding techniques which will enhance the process of data transfer on two levels. This approach involves the creation of product codes which are used on the production planning level and feature codes which are used on the operational level. “C” language computer programs have been written to create an interactive procedure which places the designer in the manufacturing loop. The impact of group technology integrated with computer-aided process planning (GT/CAPP) on the generation of a product data exchange standard (PDES) will be demonstrated.

This paper proposes a novel approach to 2D picture description. In this approach, points essential to specify pictures are defined through repetitive geometrical constructions, and the final image is drawn by referring to those points. The method fulfills the requirements for picture description: easiness, intuitiveness, and universality. In addition, to clarify the mechanism of drawing input, we formulate the specification of a drawing input system. To represent the relationships among points, lines, and circles, the specification uses the geometrical operations. We show the validity of drawing input through three applications: engineering drawing, apparel pattern-making, and Tibetan mandala image generation.

A method to find a circle, c(x,y,r), discriminates among up to eight possible solutions. Given a sufficient set of radius, centre line(s) and/or tangent line(s) the solution is the point of intersection among three planes. Including tangent circle(s) it is a piercing point of the line of two planes with a cone.

The use of graphic paintboxes is widely spread in the domain of artistic graphics. As opposed to CAD programs which are intended for engineers for the conception of objects using equations and exact values, paintboxes are builted for graphists (publicity, video, fashion…) ([WIL 84]). This distinction has lead to the introduction of the class of 2D 1/2 images. This class of images integrates notions like hidden surface elimination (according to the relative positions of objects together) without numeric data. The user has only to put the new object in relation to the other ones (for example, one can say that: “the square is in front of the circle and behind the polygon”).

In this paper we present a distributed image processing system that exploits a network of workstations, a network-based windowing system, and an object-oriented user interface. Image processing involves large amounts of both data storage and computation. Traditionally, image processing systems have been based on special purpose hardware specifically designed for imaging applications. However, the recent advances in workstation technology and networking make it possible to have the compute power and storage needed for imaging applications in a distributed environment. The network model, in conjunction with network based windowing systems, object-oriented programming techniques, and device independent image processing software, has been used to design the distributed imaging system described in this paper.

The application of a direct manipulation interface to geometric construction programs has been hindered primarily because of the “ruler and compass” approach employed in these systems. This approach limits the types of geometric tasks which can be handled, and is suitable only for a procedural interface language. It is shown that the use of a least squares adjustment for geometric computations overcomes these problems. A prototype program is presented which demonstrates the generality and usefulness of such a system.

This paper describes the application of a definitive (definition-based) programming paradigm to graphics software. The potential merits of using definitive principles for interactive graphics were considered from a theoretical perspective in [Be87]; this paper is complementary, in that it describes the insights gained through practical experience in implementing a prototype system. The main characteristics of the prototype implementation are illustrated by simple examples. Analysis of the abstract machine model underlying this implementation suggests a general purpose programming paradigm based on definitive principles that can be applied to more ambitious applications.

GRAFLOG is an experimental graphical and logical programming language. The system supports bidirectional mappings from graphical structures into logical structures, and the meaning of a complex graphical object is a regular function of the meanings of its constituent parts. Graphic scenarios are composed interactively from primitive graphical objects and relations, formally described by an underlying graphical grammar. Graphical symbols are introduced by deictic expressions, and represent individuals with their corresponding properties. Deictic expressions are used to impose semantic interpretations upon graphical symbols. Through logical expressions more complex interpretations are given to drawings. GRAFLOG is illustrated by an example in which the end-user is a cognitive psychologist. His view of the problem domain and the way he relates with the system are presented. GRAFLOG is programmed in PROLOG and GKS.

The GKS-2D standard specifications were expressed by a combination of the Entity-Relationship (ER) model with Dataflow and State Transition diagrams. Our description was not only precise and complete, but also easier to understand than the original standard, mainly because of the use of diagrams. Starting from this description a GKS system was rapidly constructed making use of a general purpose subroutine package for ER data definition and manipulation. The performance of this GKS was found to be similar to that of commercially available products. Our description was extended to GKS-3D and is generally applicable to any graphics system. We suggest that such an approach would lead to more rapid convergence in the design of graphics standards and the construction of the corresponding systems.

The emergence of conflicting and competing standards is causing confusion for users and vendors alike. The Graphics Kernel System GKS may be an approved standard, but it is not perfect. Proposals for enhanced standards would need to include additional primitives and attributes to support the modern engineering workstations, which can have sophisticated image processing and advanced 3D capabilities. Though it will take some years to define, PHIGS+ may be the way of the future.

Molecular graphics (MG) is nowadays an essential component of the basic tools used in computer-assisted chemistry. As such, it is employed in numerous applications where its role consists not only in building and visualizing chemical models, but also in simulating complex situations resulting from the dynamic properties of chemical systems. This paper presents several basic techniques used in MG by reviewing some applications developed in our laboratories in the following areas: (i) modelization of molecular architectures; (ii) real time animated representation of dynamic processes; (iii) construction and visualization of molecular properties such as electron densities and intermolecular interaction potentials.

In recent years a number of computerized graphical tools appeared on the market which are of great importance for the chemist, biologist, pharmacologist and toxicologist for the rational molecular design of bioactive compounds such as drugs. In the present paper we want to review several molecular graphics and modeling programs which can be run on microcomputers. An overview is given of available programs discussing of each its possibilités and shortcomings. In particular we introduce a new program named Waalsurf which is a high-quality molecular graphics facility which can be installed on a low-cost PC.

MOPIC (MOlecular PICtures) is a microcomputer implementation of molecular graphics software. It is designed for chemists and students, with no computer training, who need to manipulate and display molecules but do not require dedicated high performance systems. Moreover, since its graphics is patterned after vector display systems, MOPIC produces stick and dot sphere images of molecules which communicate a degree of visualization far superior to present microcomputer software.

This paper describes the graphical part of a broader project consisting of the resolution of algebraic linear and homogeneous differential equations in the complex domain.

A method is described whereby computer tomograph (CT) image slices are interpolated, an important step in the 3D visualization process. The segmented components of the images are interpolated as binary structures and then reunited. This new method of binary interpolation interpolates coordinates. The interpolation lines are bound by the image pixels. This is an element not found in current triangulation techniques.

The recent development of advanced computer graphic techniques has significantly contributed to the design of new image processing and image analysis algorithms. Most image processing workstations rely nowadays on sophisticated computer graphics for the simplification of user interactions and for the enhancement of the display of the results. In this paper we would like to outline some of the characteristics and advantages of graphic oriented user interfaces for medical image processing and analysis. Also we will review some of the new approaches in displaying complex analysis results in color coded parametric images. The use of graphics and color coded images can significantly improve the practicability of medical image analysis and allow an easier access to sophisticated quantitative algorithms for non computer-oriented clinicians.

The analysis of morphology or distribution of neurons in the brain requires sectioning of the brain which allows the cells to be viewed with a microscope, and which, moreover, is usually a requirement for the execution of the histological techniques needed to make the nerve cells visible. The section is thin relative to the size of the neuron (electron microscopy) or relative to regions of the brain (light microscopy) of interest in a particular study: Therefore, the spatial information is virtually reduced to a plane.

This paper addresses the problem of accurately modelling the motion of endoscope inside the human colon as the basis of an animated teaching aid. A problem arises from the difficulties that student endoscopists have in controlling the instrument. The human colon is a dynamically complex object capable of forming loops and changing shapes in response to the movement of the endoscope. Since the colon obviously cannot be seen from outside the inexperienced endoscopists can only rely on the TV display of the inside of the colon to deduce where the tip is. Under certain circumstances, when the colon forms into loop, no amount of pushing can advance the endoscope. A complicated manoeuvre is then necessary to straighten it to allow forward movement. At present learning the successful control of endoscopes in a colon examination is a difficult task requiring practice during hundreds of examinations. Moreover, patients can be put at risk when inexperienced endoscopists practise their skills. As a substitute for real patient, mechanical models of colon are available which are made of plastic, but they can damage the instrument. An alternative to these teaching methods is to replace them with a computer simulation. The objective would be to provide a graphical representation of the view as seen from the eyepiece of the endoscope coupled to endoscope controls similiar to an conventional endoscope. Additionally the simulator can, on command, provide views from outside the colon, readout of time elapsed, accuracy and instruction feedback to the users. The system needs to work in real time and, for wide distribution, be available on a microcomputer.

In order to aid the debugging of complex circuits, our multiple-window hierarchical object-oriented simulation environment for VLSI circuits provides powerful features for navigation in the cell hierarchy of the simulated circuit. The environment mimics this hierarchy in that each cell (instance) is treated as an object which can show its structure (static) and simulation state (dynamic signal values, probes, etc.) in its own screen window. The structural description is obtained from a circuit database. The dynamic description is accessed by interposing virtual cell simulation objects between the interface and the simulator. These objects map the possibly flattened simulation structure back onto the structural hierarchy and provide a homogeneous access for controlling the simulator and for observing the simulation results in every cell. The parallel object hierarchies of structure, dynamic descriptors and interface windows allow easy replacement of objects in a simulation at multiple levels of abstraction, and it can be adapted for a distributed simulator.

Three types of furniture parts and modelling method in FCAD system (Computer Aided Design for Furniture Structure), is introduced. Some interactive functions for modifying part models and deriving a variety of practical parts are described. Finally, the application of the modelling method to computer aided manufacturing of furniture parts is prospected.

The process of adding computerized color to black and white motion pictures and television programs has been hailed as “an ingenious technological breakthrough” and derided as “cultural butchery”. Businessmen see it as a means of breathing new life into unsaleable media properties while the original filmmakers (who do not share in the new income) see it as “vandlism” of classic works of art.

Basic concepts of urban data management and processing are introduced. Applications, types of data and operations are reviewed.Geographic queries are analysed in the context of urban data management and processing. Elementary operations are extracted, both from the access to objects and geometric computation points of view.A set of queries is designed, based on these sets of elementary operations, and measurements to be done on these queries are discussed. This constitutes a first step towards the definition of benchmarks that could be used to measure the ability and the performance of data management systems for handling and processing urban data, and graphics-oriented data in general.

GQL is a graphical formal query language for manipulating ’a database. By providing a set of pattern images, the language allows users to draw pattern graphs against the graphical schema of the database. This paper describes various aspects of GQL: the underlying model, the pattern images and graphs, and the representation of database operations.

Sail cutting is presently a trial and error procedure, as performed by most manufacturers. In this paper, a first approach to computer aided sail pattern drawing is presented. To derive the aerodynamic sections of sails circle interpolations are used, and subsequent refinements are allowed because the development and the resulting implementation are highly modular.

This is a study to determine whether the characteristics of a graphic and the position of its elements affect an individual’s efficiency in information processing. The study, which deals with bar graphs, compares the rules put forward by Tufte for constructing graphics with results obtained on multidimensional stimuli in cognitive psychology.

An ‘Integrated CCJnputer Art System’ is described which offers the ccmputer artist, illustrator and designer a much wider range of 2D manipulations than in conventional p:1int systems. The system allows non-prograrnners to explore a range of techniques based on computer geometries and algorithms, including certain types of fractals, in an interactive fashion, while still giving full control down to the pixel level. The system is under developrent, but already shows potential for generating new computer imagery not belonging to the now ‘conventional’ world of 3D photo-realism.

In this paper we briefly describe WRAP (Workcell ReAl-time Programming) which provides an integrated run-time/programming environment for a distributed robotic workcell. A robotic workcell is a distributed system consisting of a variety of elements such as multiple robots. multiple sensors and other factory machines. To synchronize and coordinate the concurrent operations of these elements WRAP uses a user defined state formalism. A sample idealized assembly application programmed under WRAP is also distussed to illustrate the programming methodology. the power and the flexibility of the system. The assembly work(ellconsists of two six degree of freedom robots. a linear stage, an overhead camera and a infrared range sensor mounted on one of the robots.

ROPSE, is a friendly unified system for both programming and robot guiding assistance. It constitutes the user interface in the industrial version of ACRO modular robots. ROPSE, implemented in Pascal on PC-AT micro computer, realizes full driving of the operator by interactive graphical means based on the tree-structured rolling menus principle and accessible by a logical mouse. Actions are selected on the screen for actual robot driving or easy and fast robot off-line programming. The system offers a set of complementary debuggIng tools: off-line graphical simulation, step-by-step robot on-line driving, and a teaching by showing module that proposes unconventional solutions.