Abstract: A method of modeling symbol sequence generation is obtained which allows the probability of symbol sequences to be estimated. The model involves treating symbols as if they are emitted as a point travels through an abstract space called a continuity map (CM), in which each position in the space has associated probabilities of emitting each of the possible symbols. This method for modeling symbol generation, combined with methods for estimating the probability of symbol sequences given smooth paths through the CM, can be applied to such problems as language modeling and anomaly/fraud detection. A fraud detection study is described that demonstrates that the invention can be used for detecting medical fraud.

Abstract: An image population having a large number of attributes is processed to form a display population with a predetermined smaller number of attributes that represent the larger number of attributes. In a particular application, the color values in an image are compressed for storage in a discrete look-up table (LUT). Color space containing the LUT color values is successively subdivided into smaller volumes until a plurality of volumes are formed, each having no more than a preselected maximum number of color values. Image pixel color values can then be rapidly placed in a volume with only a relatively few LUT values from which a nearest neighbor is selected. Image color values are assigned 8 bit pointers to their closest LUT value whereby data processing requires only the 8 bit pointer value to provide 24 bit color values from the LUT.

Abstract: A data source on a main local area network (LAN) transmits broadcast messages via bridges to remote stations. A ring buffer in a bridge on the main LAN buffers a selected number of packets received from the data source. When the main bridge on the main LAN receives a "Lost Frame Message" from a remote bridge, not on the main LAN, the main bridge backs up to the address of the lost packet in the ring buffer, and retransmits the lost packet and then continues sequential transmission of subsequent packets. If buffer is, filled above a pre-set criteria, retransmission is slowed or stopped until buffer is mostly emptied.

Abstract: An image population having a large number of attributes is processed to form a display population with a predetermined smaller number of attributes which represent the larger number of attributes. In a particular application, the color values in an image are compressed for storage in a discrete lookup table (LUT) where an 8-bit data signal is enabled to form a display of 24-bit color values. The LUT is formed in a sampling and averaging process from the image color values with no requirement to define discrete Voronoi regions for color compression. Image color values are assigned 8-bit pointers to their closest LUT value whereby data processing requires only the 8-bit pointer value to provide 24-bit color values from the LUT.

Abstract: An improved earring rack is provided and consists of a grid frame for removably holding the hook earrings thereon and two shutters, each having a plurality of small holes for removably holding post earrings with their respective friction nuts thereon. Each shutter is hinged to one side of the grid frame so that the shutters can be swung centerwards for the post earrings to be removed easily. The grid frame is mounted in a stationary vertical position to prevent the rack from falling over when the earrings are being removed therefrom.

Abstract: An air jet weaving machine in which the pick sensor and the filling stretch pipe are combined to eliminate the space in the reed normally required to locate the pick sensor. The combination pick sensor and stretch pipe are located at the end of the reed remote from the filling air nozzle adjacent the filling yarn suction member.

Abstract: An interface device called an adapter is provided for interconnecting data terminal equipment (DTE) to a communication highway. The adapter includes a plurality of command/status registers coupled through a decoding device to the data bus and control lines of a microprocessor. An Internal Modem Interface (IMI) circuit arrangement monitors the decoding device and command/status registers and depending on the contents of the command/status registers and/or the status of the signals on the interface the IMI enables/disables the receive and transmit logic functions of the adapter. The adapter also includes a circuit arrangement which monitors traffic or signals on the data highway without affecting the electrical characteristics of the traffic.

Abstract: In order to reduce data storage and transmission bandwidth requirements, a high resolution gray scale image is represented in two separate image forms, namely, a high resolution, high sampling density black and white image and a lower resolution, lower sampling density gray scale image. Generating separate black and white and gray scale image data allows the use of either image form independently. The high resolution black and white image data and the low resolution gray scale image data may be recombined subsequently to produce reconstructed high resolution gray scale image data representing the image.

Abstract: A scanner for scanning an increased area of a finger held or pressed on a concave elliptically cylindrical surface having two focal axes includes a linear light source positioned at one focal axis of a first reflective cylindrical elliptical surface, the other focal axis coinciding with one focal axis of the concave elliptical surface in contact with the finger. A second reflective, cylindrical elliptical surface has two focal axes where one of the axes coincides with the position of a linear array combination of photosensitive elements and charged coupled devices and the other coincides with the other focal axis of the concave elliptical surface in contact with the finger.

Abstract: An analog video data signal digitization and correction system scans a medium to produce video data signals representing images printed on the medium. Two reference quantities representing analog reference signals produced by scanning a densely printed image and an area of the medium having no printed image are used to establish the maximum and minimum values of a range of possible values for each video data signal. The analog video data signals are compared with this range and a corrected digital value obtained for each data signal depending on its position in the range. In generating the reference quantities each analog reference signal is converted into a digital reference signal dependent on the position of the analog signal in a range of possible values. For each analog reference signal only a portion of the range of possible values is selected in order to increase the accuracy of the corresponding digital reference signal.

Abstract: Efficient opaque metal conductors extending across the light receiving surface of a time delay and integration scanning array, are arranged at an angle to the array so as to make their interference with light, projected onto the array, substantially uniform. An anamorphic lens distorts the image projected onto the array to enable more convenient and efficient use of the array surface area.

Abstract: A two-dimensional history of reflectance values, from which black-white threshold decisions are made, is stored in a relatively small amount of memory. Updating of the history by non-linear functions is accomplished through the use of look-up tables.

Abstract: Fluorescent bulbs having two phosphors emitting different colors with different persistence are used to illuminate a target object with sequentially different predetermined mixtures of light from which color separated images can be taken.

Abstract: A CCD imaging array is operated in a time delay and integration mode to generate an array of charge packets corresponding to the light intensity at the picture elements of an irradiating image. The charge packets from the imaging array are gated in parallel to a line image storage array that stores n rows of charge packets in a columnar relation. A replicator circuit generates a replicated sum charge for each column of charge packets of the line image storage array. Each replicated sum charge corresponds to the sum of the n charges stored in a column at a particular instant in time and the charge stored at the (n+1)/2 position of each column is the middle charge for the associated replicated sum charge. Parallel delay gating electrodes gate a row of replicated sum charges to an area average serial shift register at the same time that a row of corresponding middle charges is gated into a focused element serial shift register.

Abstract: A color facsimile transmission apparatus wherein a beam of white light is dispersed into component colors by a prism. The color dispersed bands of illumination are focused to irradiate a portion of a moving document. The reflected, colored light from the document is focused on the surface of three charge coupled devices (CCDs) so that the red illuminating band irradiates a first CCD and the green and blue bands respectively irradiate the second and third CCDs. The movement of the document causes each line of the document to scroll through the illuminating band over each of the CCDs. The CCDs are operated in the time delay and integration mode to generate charge packets corresponding to the picture elements of the red, green and blue component line images of each line of the document.

Abstract: A circuit is disclosed for obtaining a quantity of electrical charge carriers equal to and representative of the algebraic difference between the values of two original quantities of electrical charge carriers. The circuit includes two charge coupled device input shift registers. Each of the two shift registers incorporates at least three potential well electrodes and interposed transfer electrodes such that charge packets can be shifted from well to well. One potential well in each of the two shift registers is operated in a floating gate electrode mode, and the two potential wells are connected at a common node which services as the input to a third or output charge coupled device shift register.

Abstract: A document scanner array which includes means for compensating for temporal changes in a light source, filters in the optical path, changes in the lens aperture and the like. The scanner array is comprised of a charge transfer device chip, which includes an exposure control section and a document information receiving section. A document holding means includes a reference indicia, with a reflection of the image of the reference indicia being imaged on the exposure control section simultaneous with the reflection of the document being imaged on the document information receiving section, in response to illumination from the light source. A common control signal concurrently applied to each section determines the time the reflected images are processed. The common control signal is terminated in response to the exposure control section sensing a given amount of light from the reflected image of the reference indicia.

Abstract: An image scanning system which alternately illuminates a charge transfer device with a focused, then defocused image of a line at a time of a document in order to provide comparison signals for binarization of the image concurrent with cancellation of dark current and sensitivity variations. The system includes focus-defocus, variable optical path length optical elements which are operated at a reduced speed by processing a line array at a time instead of a single picture element at a time in the image. A single charge transfer device chip comprises the scanner array, and includes a simple analog line memory comprised of light responsive elements, shift registers, gates, and a 1-bit comparator to perform the binarization. The charge transfer devices may be comprised of charge coupled devices (CCD) or bucket brigade devices (BBD) building blocks.

Abstract: A charge transfer device transversal filter is provided having electronically controllable weighting factors. Two equivalent and corresponding charge transfer device shift registers have floating electrodes sandwiched between the sensing electrodes and the substrate for biasing the underlying depletion regions and thereby controlling the effective capacitances thereof. Depletion capacitances are controlled so as to produce desired weighting factors when corresponding and otherwise equal charging currents are substrated one from the other.

Abstract: A holographic laser beam deflector has an approximately flat field scan. The scan is produced by illuminating a rotating hologram on a circular surface by a laser beam. The virtual source of the illuminating beam is offset from the center of rotation of the hologram. The hologram is produced by interference of a pair of coherent laser beams, each beam having a source or virtual source on a line containing the rotational center of the surface of the hologram, one source or virtual source outside the surface, the other inside. The virtual source for the inside beam is also displaced from the center of rotation.