Norio Tayama

A speedy voxel‐tracing method for cutting solid display of 3‐D images

This paper describes a high-speed image-generation method which arbitrarily cuts out a three-dimensional (3-D) digital object (3-D image) in a 3-D frame memory and three-dimensionally displays its internal shape realistically. The idea of ray-tracing is applied to a 3-D memory space so that a high-speed ray-tracing is carried out in it by using a 3-D digital differential analyzer (DDA) called the reversible step type. To obtain cross points of a ray (straight line) and an object in a 3-D frame memory, an algorithm indicating address orders of all the voxels through which the ray passes is shown. A viewpoint of obtaining a realistic 3-D representation with a smooth surface from an irregular digital surface, calculation method of normal vector is important. A new factor, “existential density,” which can be obtained by a simple calculation around the surface of the digital object, is introduced; and a method to obtain a normal vector from this factor using the distribution of density gradient also is described. Since this calculation requires 3 × 3 × 3 operators, addition/ subtraction of integers and shift operations, the operation is high speed and suitable for hardware. The proposed algorithm is confirmed by a computer simulation which produces 3-D images with smooth shading.

A neural network structure based on pulse-train arithmetic

A Hopfield neural network with a simple structure based on a pulse-train signal arithmetic has been developed. It is both analog and digital in nature and involves hybrid system concepts. It is implemented by using only two types of principle elements, a neuron module and a synapse module. The neural network proposed performs pulse-train signal processing directly. The advantages of this design are a simple operating mechanism, homogeneous implementation and expandability.<<ETX>>

A simple neural network structure with good expansibility based on pulse-train arithmetic

The authors discuss the simplification, expansibility, and modularity of a neural network architecture based on pulse-train arithmetic. The simplified architecture is implemented by connecting basic modules in parallel with a single bus line of a control processor. The basic module is composed of a neuron unit and synapse unit. Thus, the expansion of the network can be easily realized by increasing the number of basic modules. The basic operations of this network architecture have been confirmed by computer simulation of the traveling-salesman problem.<<ETX>>

Development of interactive multimedia database

This paper presents interactive multimedia database (IMDB) which makes image processing possible on World Wide Web (WWW). It is different from typical image database to provide only storage and retrieval function of image data. The higher performance of electronic parts such as a transistor, the inner structure becomes more complicated. Non-destructive evaluations of the high-precision electronic parts are necessary to maintain the quality. As the first approach, we visualize inside state of electronic parts as two or three-dimensional images with x-ray CT (computerized tomography), and we aim at the efficiency of non-destructive test by registering these images with database. IMDB provides cross section of electronic parts, and three- dimensional visualization from various angles, and animation to support the test of defects. The point we want to make is that this database is different from current client/server type database that retrieval function is main function, and IMDB enables users to process image such as edge extraction on WWW network. In addition to CT image, IMDB receives various images filmed with digital camera and image scanner from users, and executes the request of image processing and show the results.

Design theory of three-dimensional memory system for image processing

This paper considers the parallel processing system for a three-dimensional (3-D) image, and presents a theory for constructing the 3-D image memory system with a parallel access function. Until now, studies have been made on the condition for the feasibility of parallel access for the assignment function (parallel access function) of the individual module. In the view of the authors, the traditional parallel access function cannot be used in the general theory of the memory module assignment. A new concept called “memory module assignment permutation” is proposed. This paper considers the following problem based on the proposed concept. A 3-D image memory system with seven kinds of parallel access functions is to be realized, and the problem is to clarify the mathematical condition for the memory module assignment permutation. The problem is formulated and the design condition is stated. In other words, the necessary and sufficient condition is derived for the memory module assignment permutation so that the parallel access can be made for the three kinds of line-mode (X, Y, and Z axis), three kinds of plane-mode (perpendicular to the axis), and the cubic-mode.

The Design Theory of Two‐Dimensional Memory Systems for Image Processing

This paper considers the two-dimensional image memory system with the parallel access function required in the parallel image processing, and presents the design theory for its construction. So far, numerous studies have been made on the condition for the parallel access for individual memory module assignment expression. The authors believe, however, that the general theory cannot be developed by the earlier functional expression for the memory module assignment. This paper presents a new concept of “memory module assignment permutation.” Based on the proposed idea, a problem is considered concerning the mathematical properties of the memory module assignment permutation to realize the two-dimensional image memory system with various kinds of parallel access. The problem is formulated and the design condition is described. In other words, the horizontal and vertical line mode as well as the plane-mode are considered. The necessary and sufficient condition is derived for the memory module assignment permutation to realize the various parallel access.

Distance transformation and reconstructable skeleton of three‐dimensional digital binary images

This paper considers the filled three-dimensional image, such as the one obtained by superposing several CT images, and discusses the distance transformation and the skeleton for the three-dimensional filled object. From the viewpoint that the internal structure of the three-dimensional object should easily be understood, there has been devised a distance transformation, where the distances from all points inside of the three-dimensional object to the surface are determined. There has not been presented, however, a method which is applicable to the general situation. From such a viewpoint, this paper introduces a new notion of fundamental neighbor set, and proposes a general parallel and serial algorithms for the distance transformation. As an example, the notion of the reconstruetable skeleton for the three-dimensional object is proposed, and the parallel and serial algorithms are presented, which can reconstruct accurately the original three-dimensional object from the skeleton. The reconstructable skeleton will be useful as a means of information compression for the three-dimensional image, which requires especially a large amount of data for the representation. The validity of the proposed algorithms is verified by computer simulation.