Abstract: In an image extraction system, an extracting part for extracting wide lines, an extracting part for extracting narrow lines and a frame detector detect a frame from a pattern which is extracted by a connected pattern extracting part. An attribute adder adds attributes of a character (graphic and symbol inclusive), frame, and a contact pattern of the character and frame to a partial pattern, and a separating part separates the frame from the contact pattern. An intersection calculator calculates intersections of the character and frame, and the calculated intersections are associated by an intersection associating part. An interpolator obtains a character region within the frame and interpolates this region based on the associated intersections. A connection confirming part confirms a connection of the pattern with respect to the extracted character pattern, and patterns confirmed of their connection are integrated in a connected pattern integrating part to thereby extract the character.

Abstract: A reading apparatus serving as an input means for simultaneously reading printed image information from both the front and back of a document, converting the read information into digital data, and storing the digital data on a data recording medium, such as an optical disk, comprises document carrier mechanisms, a front reading sensor, a back reading sensor, a front size determining sensor, a back size determining sensor, a document size comparator, and a reading control circuit. These components are incorporated in the reading apparatus. The document carrier mechanisms carry documents inserted through a slot through the apparatus and eject them through an outlet. The front data reading sensor and back data reading sensor read printed images from the front and back of a document. The front size determining sensor and back size determining sensor are located upstream of the front data reader in the document carrier path, and detect the sizes of the front and back of a document.

Abstract: A smoothing technique for processing digital images is presented. The smoothing technique of this invention comprises a two-dimensional digital hysteresis filter which utilizes a "two-dimensional automatically adjusting variable mask" for intensity processing. As a result, the filter is independent of the image size and content and cannot alter the size of any significant intensity features, even if they are as small as one pixel in size. The intensity processing technique provides the basis for a separation of the image information into basic information groups. The two-dimensional hysteresis smoothing technique of this invention calculates the intensity value of smoothed pixels using a set of one-dimensional hysteresis lines at various angles running through each pixel in an image. Each one-dimensional hysteresis line calculates a new intensity value for each pixel in an image.

Abstract: A method of arbitrary image resolution conversion takes an original print ready image at a first resolution and orientation and simulates printing of that image by creating a representation of a page at a particular resolution and orientation using a suitable spot model for the reconstruction function. The image is then filtered with a low pass filter and sampled with an aperture that corresponds to the desired output. The image data, which may be definable at several bits per pixel, is then made print-ready by reducing the number of bits defining each pixel to a valid output state. The reduction step is accomplished through error diffusion methods that maintain the local area gray density level. The filtering step reduces moire in the converted image.

Abstract: An image processing system which processes image data on a line by line basis. Edges in the images are detected as discontinuities in intensity. The intensity profile between detected edges is determined and matched to a polynomial function. Data representing the detected edges and data relating to the detected profile are stored as a data set for further processing. The data set for one image may be compared to that for another similar image in order to match features between images and thereby produce three-dimensional information about a scene.

Abstract: An improved method for providing a more faithful and pleasing rendition of a pixel pattern is described herein. This method first places the original pixel pattern in a bit mapped memory. A small portion of this bit mapped pattern, called an input tile, acts as an index for a look-up table containing a corrected pixel pattern portion (an output tile) corresponding to the target pixels to be corrected in the input tile. This process may also be performed using an algorithm. To derive the output tiles, a high resolution bit map is created from an input tile. A triangular portion of each convex corner in the high resolution bit map is chopped off, and a triangular portion of each concave corner in the high resolution bit map is filled in. The resulting high resolution bit map is then filtered using a low pass filter. The size of these triangular portions and the extent of filtering are determined using a function based on the vulnerability of pixels to being eliminated during the above smoothing process.

Abstract: Images such as product images are identified as substantially similar to one or more reference images by finding pixels which have substantially the same value or values in all or substantially all of the reference images. Image erosion and/or dilation may be used in processing the reference image data to help identify pixels which can always be expected to have the same value in all acceptable images. The values associated with the corresponding pixels in a product image are combined and compared to an expected value. The values associated with pixels which correspond to regions not similar in substantially all of the reference images are combined to produce a discriminant function. The product image is identified as substantially similar to the reference image or images only if the combined values from the product image compare favorably with the expected value, and the dot product between the product image and the discriminant function is equal to a predetermined value.

Abstract: A method and associated apparatus for enhancing the display or print of greys in a two-tone digitized image consisting of storing a portion of a binary image map including a central bit under examination to be displayed and in testing if the said central bit is representative of an isolated dot or whether in a region around the said central bit in one of a plurality of predetermined directions there is an isolated dot at a predetermined distance from the said central bit, to display the central bit as a dot of reduced dimensions in such a way as to convert the display of an isolated dot in the display into a plurality of smaller dots with aggregate areas equivalent to the isolated dot and distributed around the location of the isolated dot identified from the image map.

Abstract: An image transforming apparatus comprising a means for holding an input image signal, a means for holding geometric parameters for the image signal in a photographing mode, a processing means for geometrically transforming the image signal by the use of such geometric parameters, and a means for outputting the image signal transformed geometrically by the processing means. There is also provided a method of transforming an input image by employing the above apparatus to realize a tilting process and a depth-effect emphasizing process for the image obtained from a video camera, an electronic still camera or the like.

Abstract: An integrated bulk food sorting and analysis apparatus comprises a wide-belt product conveyor which receives and conveys a laterally-distributed stream of bulk food articles past a product diverter. An upstream camera is positioned to view the stream of food articles upstream of the product diverter. Automated sorting logic is responsive to the upstream camera to individually determine optical characteristics of each food article. The product diverter is responsive to the automated sorting logic to divert a portion of the food articles from the stream depending upon their individual optical characteristics. The apparatus also includes a downstream camera positioned to view the stream of food articles downstream of the product diverter. A data processor is responsive to both the upstream camera and the downstream camera to periodically examine samples of food articles and to calculate upstream and downstream quality statistics regarding the stream of food articles.

Abstract: A novel automatic inspection technique and apparatus is disclosed for differentiating differently colored regions of an object or surface (such as solid state wafers, printed circuit boards, or more general applications for distinguishing regions of one color from their neighboring or surrounding areas of different color), involving the use of at least a pair of separate detectors of light images (CCD cameras preferably) reflected from the object or surface and provided with optical filters of different colors corresponding to the differently colored regions of the object or surface; and an electronic filtering processing by multiplying the detector signals by different weighting coefficients selected to maximize the ratio between the filter responses to optimize signal contrast, and then linearly summing the multiplexed signals. Independence from wide light intensity variations and in color variations in the production or processing of the object is thereby obtained.

Abstract: In digital data conversion apparatus and method, class data are generated in association with reference interpolated data for each of a plurality of classes on the basis of a reference high definition digital video signal which includes a reference standard definition digital video signal in addition to the reference interpolated data. The class data is stored at respective addresses in a memory. A standard definition digital video signal representing pixel values is received and then clustered so as to produce a class corresponding to the pixel values of the standard definition digital video signal. The class data is retrieved from the memory address which corresponds to the class of the standard definition digital video signal, and interpolated data is generated in accordance with the standard definition digital video signal and the retrieved class data with such interpolated data constituting a high definition digital video signal.

Abstract: A method for estimating the shape of a body that emits electromagnetic energy using a sensor (10) having an image plane (15) and the ability to measure the intensity of the electromagnetic radiation (12) received at a point on image plane (15) comprises the steps of first expressing the intensity of radiation (12) reaching image plane (15) as a function of surface gradient parameters (p,q) for a pre-determined point (P.sub.i) on object (14). Next, the method requires measuring the intensity (E(s,t)) of radiation (12) that sensor (10) senses at the point (P.sub.i) on image plane (15). The method then requires determining the values of the surface gradient parameters (p,q) for the point (P.sub.i) on object (14) that minimizes the difference between the expressed intensity (F.sub.1 (p,q) for the pre-determined point (P.sub.i) on object (14) and the measured intensity (E(s,t)) at pixel (p.sub.i) on image plane (15).

Abstract: A method of stably identifying and characterizing a transient contained in a sampled set of data, the data representing a physical event, where alignment of the sample points with the transient is typically arbitrary. The method includes the step of analyzing the first data set using a predefined mathematical operator to produce a second data set, detecting and isolating the discrete data pattern in the second data set, determining a location of the detected and isolated data pattern in the second data set, measuring a difference between the location of the detected and isolated data pattern and a predetermined location of the predefined mathematical operator, adjusting preselected parameters of the mathematical operator in accordance with the measured location difference such that the adjusted mathematical operator is aligned with the detected and isolated data pattern, and reanalyzing the first data set using the aligned mathematical operator.

Abstract: The invention relates to a stylus-based user interface for computers. It describes a process for separating a stylus-based application program from the procedures used to implement stylus-based, user driven error correction processes. This separation allows error correction procedures to be used by many applications, providing consistency in the user interface and saving application development costs through reuse of these procedures.

Abstract: A scanning system for orienting bar codes use a first stage made up of a low-resolution scanner and a digital signal processor (DSP), and a second stage made up of a high-resolution scanner and a decoder. The first stage scans the surface of an object to obtain an image, which the DSP analyzes to determine the location and orientation of any bar codes. The DSP does this by dividing the image into windows and finding the windows which contain part of a bar code. The second stage receives the results from the DSP and scans the bar code on the object.

Abstract: A method of operating an image recognition system including providing a neural network including a plurality of input neurons, a plurality of output neurons and an interconnection weight matrix; providing a display including an indicator; initializing the indicator to an initialized state; obtaining an image of a structure; digitizing the image so as to obtain a plurality of input intensity cells and define an input object space; transforming the input object space to a feature vector including a set of n scale-, position- and rotation- invariant feature signals, where n is a positive integer not greater than the plurality of input neurons, by extracting the set of n scale-, position- and rotation-invariant feature signals from the input object space according to a set of relationships I.sub.k =.intg..sub..OMEGA. .intg.I(x,y)h[k,I(x,y)]dxdy, where I.sub.

Abstract: The fill processing time for filling a closed figure identified by one stroke is reduced. The difference between the total number of line segments in the plus Y axis direction and that in the minus Y direction intersecting a line perpendicular to a specific first direction is counted when a figure identified by one stroke is scanned in the X axis direction. Pixel positions in a range in which the counting result is not "0" are filled.

Abstract: A field test apparatus and method is disclosed in which gradual movement of the head supporting chin cup is used to maintain a centered relation between the eye being tested and the trial lens frame holding the required prescription for optimum vision of the patient. During the test, alternating illumination is provided in the infrared between a central corneal reflection and an iris illuminating source which imparts a high contrast bright circular iris image surrounding a central dark pupil. Content addressable memories are used to delineate areas of contrast change for rapid and abbreviated microprocessor analysis of conventional RAM images. Both the corneal reflection image--which is a brightly illuminated spot on an otherwise dark background--and the pupillary image--which is a dark pupil surrounded by a brightly illuminated high contrast image of the iris are approximately located at their boundaries using content addressable memory.

Abstract: An image recognition device includes an input section for inputting an image; a pattern extraction section for extracting a plurality of predetermined patterns from the input image; a first feature extraction section for extracting a first feature indicating the direction in which the first pattern of the extracted predetermined patterns exists with respect to a point on a second pattern of the extracted predetermined patterns; a second feature extraction section for extracting a second feature indicating an amount of the first pattern existing in the direction with respect to the second pattern, based on the first feature; and a recognition section for recognizing the input image based on the second feature.