Abstract: An image processing system has an input portion to which pixel data is inputted, a convolution processing portion processing each set of pixels in a predetermined convolution area so as to output a processed data for every predetermined convolution area. A computational portion calculates a parameter necessary for image processing according to the processed data from the convolution processing portion. A converting portion includes a selecting means, to which the output of the computational portion is inputted, a high speed memory to which the output of the selecting means is inputted and a light computational portion for performing light computation to which the output of the high speed memory is inputted and the output of which is inputted to the selecting means.

Abstract: An image thinning method for obtaining a resultant image of one pixel width without removing the object being thinned when the object changes width from 3 pixels to 2 pixels. An index pixel value is defined to be a value that is not used in the image. The object is scanned so that each pixel is examined sequentially. The pixel value of the pixel being examined is changed according to the defined rules so as to create a thinned image with a width of one pixel.

Abstract: An image verification system for verifying whether a sample image corresponds to a reference image registered beforehand. A relationship between the sample and the reference image is stored when the sample image corresponds to the reference image. The sample image is compared with the reference image by verifying the sample one and the reference according to the relationship between characteristic values of the sample and reference.

Abstract: A data processing system comprising a plurality of neural layers characterized in that a part of neurons in one of the layers are connected to neurons in the following layer.

Abstract: An analog circuit which performs weighting and thresholding for a neural network. Each neuron of the neural network includes an operational amplifier receiving an input signal, the output of which is connected to a transistor. The transistor conducts only when the output exceeds a predetermined value thereby providing a threshold function. The output of the transistor is connected by a variable resistance to other inputs of other neurons. The output of each operational amplifier thereby corresponds to a weighted version of the input signal, which is adjusted for threshold and is also dependent on other neurons of the network.

Abstract: The present invention makes it possible for unskilled to inspect the total surface of a liquid crystal panel accurately in short time. It is defined that a liquid crystal panel can be divided into a single part which is a constituent, isolable, and an inspectable area, and that the single part to be "unit area of image" which is defined as a pattern to be inspected. Before the inspection, a unit area of an image without defect is selected from a liquid crystal panel to be inspected. The upper limit reference pattern and the lower limit reference pattern are generated by giving the maximal brightness in a convolution and adding the predetermined brightness and by giving the minimal brightness in the convolution and subtracting the predetermined brightness, respectively, to each pixel of the convolution.

Abstract: It is possible to inspect a printed matter precisely because lines are surely and accurately extracted by the present invention. Extraction of lines are the important point for the judgment if a character is correctly printed or not. The printed matter inspection method according to the present invention comprises the following steps: i) an image of a predetermined area including a line is input; ii) a density projection is performed on said input image; iii) a start point and end point of change of density are emphasized in said density projection; iv) the center pixel in each predetermined convolution is substituted by the minimal density pixel in each said convolution; v) a density projection is performed; and vi) said printed matter defective or not is judged according to said density projection in v).

Abstract: A number plate recognition system has structure for inputting an image; structure for calculating outlines of each configuration included in an input image, as well as for recognizing a configuration with the maximal outline size as the plate area. The areas in an image except the above plate area are masked, and the numbers are determined by using Euler numbers after calculating Euler numbers for each configuration within the above plate area.

Abstract: An image processing system has an input portion to which pixel data is inputted, a convolution processing portion processing each set of pixels in a predetermined convolution area so as to output a processed data for every predetermined convolution area. A computational portion calculates a parameter necessary for a image processing according to the processed data from the convolution processing portion. A converting portion includes a selecting means, to which the output of the computational portion is inputted, a high speed memory to which the output of the selecting means is inputted and a light computational portion for performing light computation to which the output of the high speed memory is inputted and the output of which is inputted to the selecting means.

Abstract: An image processing system has an input portion to which pixel data is input. A convolution processing portion processes each set of pixels in a predetermined convolution area and outputs a processed data by every predetermined convolution area. A computational portion calculates a parameter necessary for a image processing according to the processed data from the convolution processing portion. A converting portion comprises a selector receiving the output of the computational portion, a high speed memory receiving the output of the selector and a light computational portion for performing a light computation to which the output of the high speed memory is input and the output of which is input to one input of the selector.

Abstract: A method for comparing two images including ridges with each other, comprising steps from i) to iv).i) Calculating density value of each pixel of an image to be examined and reference image; ii) detecting a direction with minimal change between each pixel and adjacent one, concerning to each pixel; iii) dividing the image to be examined and reference image, calculating an index according to the direction with minimal change as to each block; iv) using the index, it is judged if the image to be examined and reference image substantially coincide to each other.

Abstract: A system for determining a velocity of fluid flowing around an object generates a signal for an image of a fluid-flow surrounding the object. Based on the signal, the system determines a centroid for each particle in the fluid-flow, and calculates distances between centroids of two nearest particles. The system then determines a velocity distribution of the fluid-flow based on the distances.

Abstract: A 3-dimensional configuration recognition system comprising: a means for projecting a checker pattern onto a surface of a solid object; a means for thinning said checker pattern and for generating a thinned checker pattern; a means for extracting cross points of said thinned checker pattern; a couple of means for detecting 2-dimensional coordinate of said cross points; and a means for calculating 3-dimensional coordinate of said cross points from said 2-dimensional coordinate and positions of said couple of means.

Abstract: A method for storing information defining a configuration in an image scans an image, and distinguishes a background density level from a density level of the configuration. For scan lines, start pixels are identified as the first of a contiguous group of pixels having the configuration density encountered in the direction of a scan, and end pixels are identified as the last of a contiguous group of pixels having the configuration density. Chain codes are generated linking start and end pixels among differing scan lines, and start and end pixels are also indexed. Chain codes and indexes are stored as image representations.

Abstract: A method for processing images is useful for generating a chain code representation of a configuration of pixels in an image. The method stores a pixel representation in first frame memory. For each pixel, "convolution information" is stored in a second frame memory. Convolution information indicates whether neighborhood pixels are part of the configuration. In one embodiment, convolution information for an object pixel is a bit string. The location of a bit in the string corresponds to a direction of displacement from the object pixel. The value of a bit in the string indicates whether a neighborhood pixel located in the corresponding direction is part of the configuration. A first method for generating a chain code expression examines the convolution information for a bit which identifies an adjacent configuration pixel. Examination begins from a bit location determined from a prior code in the chain. The position of the identified bit is then the next code in the chain.

Abstract: An inspection method of liquid crystal panel by comparing characteristics of histogram of a part mounted on the panel with a reference histogram which is obtained from a part of the panel to be inspected.

Abstract: A method of generating a binarized image from an original image. Error values of binarized pixels in the neighborhood of a pixel to be processed are used to modify the pixel to be processed. The error values are filtered according to weighting factors which vary as a function of the density value of the pixel to be processed independent from the density values of the pixels in the neighborhood. The weighted error values are used to modify the pixel to be processed prior to comparing the pixel to be processed to a threshold.

Abstract: A fingerprint photographing system comprising: a prism having a detection surface where a finger is pressed thereon; a luminous source for irradiating light to the detection surface; a camera for detecting the light reflected upon the detection surface; and means for comparing the image before and after the finger is pressed upon the detection surface and for storing the image of fingerprint to a memory when the number of pixels of the image with a predetermined density difference is in the range of predetermined number.

Abstract: A method and apparatus for aligning first and second images uses closed contours of pixels from the images. In the preferred embodiment, a series of pixels is selected from along a contour in each image. The contours can be circles. The two pixel series are compared to determine a degree of concurrence. One series is shifted by one pixel relative to the other, and concurrence is determined after the shift. The series are shifted repeatedly until all shifts have been measured. The shift distance which results in the maximum concurrence is selected as a measure of the amount of rotation needed to align the images. After rotational alignment, one image is displaced by one relative to the other and the displaced images are compared to determine a degree of coincidence. The images are displaced eight times, each displacement testing one position in a different direction. When a displacement is found to give a higher degree of coincidence between the images, the images are displaced in that direction.

Abstract: An image verification system for verifying whether a sample image corresponds to a reference image registered beforehand. A relationship between the sample and the reference image is stored when the sample image corresponds to the reference image. The sample image is compared with the reference image by verifying the sample and the reference according to the relationship between characteristics values of the sample and reference.