Abstract: A method for extracting a three-dimensional color vector from a three-dimensional color space in which correlated features are distributed includes steps of projecting the three-dimensional color space on a two-dimensional plane, displaying the two-dimensional plane on a display screen of a display, designating a line segment on the display screen by an input device, and executing the above steps at least two times by changing the angle of projection between the three-dimensional space and the two-dimensional plane to extract the color vector.

Abstract: A function of simulating change of weather and time in one scene by changing a sky area and an object area in a natural image while maintaining a relationship therebetween. The natural image is separated into a sky area and an object area excluding the sky area. The color of the object area is generated from a reflection model of the object, based on the color of sunlight, the color of a sky light, and a reflectance of the object. The sky light is determined by averaging pixel values of the sky area. The color of the sky light can be determined based on the color of the sunlight and meteorological conditions such as clouds and fog. The color of the sunlight, the meteorological conditions and the color of the sky light can be determined by specifying parameters representing time and weather. By changing the parameters, a simulation for changing the time and the weather can be performed on the entire natural image.

Abstract: A first image is inputted, and an object area for mapping is determined on the first image. Transformation parameters for the mapping are calculated pixel by pixel from the attributes of the first image in the object area, a second image to be mapped into at least one part of the object area is generated, and the second image is mapped pixel by pixel in the object area of the first image operating the transformation parameters so as to generate a third image. When an image simulation for changing the color and texture in an area of a color or monochromatic gray scale image is performed, an image simulation for ignoring effects of fine texture on the initial image and reflecting the shadow and shade can be realized.

Abstract: An information reference method includes a computer, a storage unit for storing display information to be processed by the computer, a display unit for displaying display information stored in the storage unit, and an indication unit for indicating any desired point on the screen of the display unit. The information reference method comprises a step for displaying a portion of display information stored in the storage unit, a step for indicating any desired point of a display image on the screen by the indication unit, and a step for selecting one of a plurality of display information linked in advance to a displayed image. The information retrieval method further comprises a step for providing, within display information, link information for showing the degree of relevancy between a displayed image and the plurality of display information, and a step for mutually comparing link information at an indicated point to decide display information to be selected.

Abstract: An image data compression method is disclosed, in which a sheet of image is divided into a plurality of partial images of small area, and in reconstructing an image, the fidelity to the original image is changed in accordance with the utility of the information contained in each partial image, and the image data is compressed in accordance with the particular fidelity.

Abstract: An image simulation method and system which is capable of preparing images in a simple way by running image editing tools without programming by a compiler language. The image simulation system includes a computer system for preparing an embedded image in a way that parts of a digital image or the image textures or patterns are embedded in any of several digital background images, with its geometrical shape deformed, or with color data changed as needed, or for preparing an image, color data of which are changed at particular portions thereof. The system further includes circuitry for changing the geometrical shape in preparing the embedded image, for changing the color data, for designating embedding positions, and for storing selection data for the image to be embedded, selection data for the background image, and control data for the images and such processes, and for reproducing the images on the basis of the data set up in advance.

Abstract: A texture mapping method projects a specified texture picture onto the surface of an object in another specified picture. Three-dimensional data on the surface shape of the object to be projected is not known previously. The three-dimensional data on the surface shape of the object is estimated on the basis of the data in the objective picture to attain the texture mapping.

Abstract: In order to accomplish a color adjustment capable of determining a color at a high degree of freedom, a variety of color expressions and corresponding input instructions are correlated to reflect the instruction of a color change by a certain expression upon another expression. In order to accomplish a color change considering the changes in the ambience, moreover, each vector of an object color, a light source color, a color texture, an ambient color and an input bias is decided to determine the separated component values of the former three vectors of a picture element value so that the changed color of the picture element value may be determined after the vector change from the separated component values and the changed vectors.

Abstract: The king of a partial image designated on the display screen is determined and desired processing is effected at high speeds. For this purpose, the region designated on the displayed image and the attribute or identity data of that region are stored corresponding to each other. The attribute or identity of the partial image corresponding to the region is retrieved by using the attribute or identity data. The region of the partial image having a predetermined attribute inclusive of the retrieved attribute is picked up, and a predetermined operation is effected on the image data in the region that is picked up.

Abstract: By using the state transition of a highly interconnected neural network, in order to solve a combination problem, an energy function is set by the following procedure: (i) the energy function is set in correspondence to the size of the combination problem; (ii) the energy function is set for a combination problem to be solved by using an energy function which solved another combination problem of a different size from the combination problem to be solved. Also, in order to solve a problem involving the cutting out a specific image from a whole image, as a combination problem when obtaining pixels representing a contour of an object, the energy function is set by either (i) or (ii) above.

Abstract: An image processing area is divided into a plurality of small areas which are allowed to overlap one another. One or a plurality of pixels in a region to be segmented and in a region not to be segmented are specified. A representative point of the region to be segmented and a reference point of the region not to be segmented are determined from the color information of each of the pixels. A discriminating plane, a plane or a curved surface that intersects the line segment connecting the representative point and the reference point a right angles is used in processing. By the processing a particular region can be segmented out of the image. Further, a shape model is made by two-dimensionally projecting from a given direction the three-dimensional structure expressed by a line drawing consisting of nodes and curves connecting the nodes based on priori property relating the structure of the region to be segmented in the image.

Abstract: Method and system of selecting desirable time series data pieces from many time series data pieces. Characteristics extracted from time series data pieces are applied to neurons and neuron state fluctuations are supervised and evaluated through execution of neurocomputing to select desirable time series data.

Abstract: In order to highly efficiently compress images of large quantity obtained by the photographing from a plurality of view points, provision is made of means for photographing a three-dimensional body or the like from a plurality of view points, means for forming a correlated image having a high correlation relative to the neighboring images by effecting mean processing for the neighboring images that are photographed, means for constructing, in the form of tree structure data, a set of the photographed images and the images formed by the mean processing, means for estimating the image of a lower rank by the one rank higher image in the tree structure data, and means for encoding an estimated error.

Abstract: The present invention is capable of registering and reading out a logical element for which the state of the output pin changes. The system includes an input side reading out circuit for reading out the kind of logical element and the states of all the input pins thereof, a decision circuit for deciding the presence of the output pin that the status change is produced on when a logical operation is carried out according to the kind of logical element, an output side reading out circuit for reading out the information related to the logical element of the output pin producing the status change, and an exchange sending circuit for sending each information read out from the output side reading out circuit to the desired registering and reading out circuit for precise high speed logic simulation of a large scale logic circuit containing MOS-type logical elements.

Abstract: Numerical data in a two-dimensional distribution for encoding picture data or voice data at high efficiency is divided into blocks being not equal in length depending on an amount of variation of each numeric data. The numeric data within the divided blocks is converted into data corresponding to an apex of each block, and the converted data is encoded into tree-structure data. Image data is divided into blocks being not equal in length depending on amount of variation of brightness, and the brightness within the divided blocks is converted into data corresponding to an apex of each block, and the converted data is encoded into tree-structure data. Range of values to be taken by data is divided into a range of normal values and a range of abnormal values, and a short bit number is assigned to the quantizing level in the range of normal values and a long bit number is assigned to quantizing level of other data, thereby the variable length coding at two steps is performed.

Abstract: To efficiently compress image data, the image data is divided into rectangular blocks of non-equal length in a hierarchal structure in accordance with a brightness change, and is encoded into tree-structure data for each of the divided blocks so that the mean distortion of the image data inside each block does not exceed an allowable value.

Abstract: A processing method and apparatus for providing high speed computations including an exchange network for discriminating a first processing node which can execute a predetermined arithmetic operation. The processing node being capable of executing the arithmetic operation and determining a second processing node to which the output of the executed node is to be applied. The invention may include a plurality of first and second processing nodes in parallel to enable high speed computations.

Abstract: In a method of estimating a fracture point of a pipe line network, a group of estimating rules based on distribution characteristics of extraordinary pressure variation, pipe line characteristics and flow continuity requirements in a fracture accident monitoring system for such networks represented by a large scale water distribution pipe line network are applied to reference data of flow rates and pressures both obtained in the normal consumption condition, and to measured data of the flow rate and the pressure to estimate the fracture point. In the method, application order of individual ones of the group of the estimating rules is decided in accordance with data conditions of the pipe line network.

Abstract: Method and system for controlling fluid transportation in a pipe network are disclosed. Pressures and flow rates at a predetermined number of points in the pipe network which number is smaller than the number of demand nodes in the pipe network are detected, demand volumes at the demand nodes are estimated based on known flow and pressure balance requirements for the pipe network and the detected flow rates and pressures, and the flow transportation in the pipe network is controlled in accordance with the estimated demand volumes such that the pressures at the demand nodes are brought to a reference pressure so that the pressures at the demand nodes at points other than the predetermined points are also brought to the reference pressure.

Abstract: A control method and apparatus incorporate a network model simulating hydraulic phenomena within a water distribution network whereby the pressures and flow rates within the water distribution network are controlled by an optimizing unit for inputting the current total demand as a model input and computing manipulated variables or control amounts for optimizing the pressures and flow rates in accordance with a network algorithm and a correction computing unit for computing the necessary control amount correction values for correcting the deviations from the estimated pressures and flow rates in accordance with a sensitivity matrix indicating the effects of variations in the control amounts on the pressures and flow rates.