Abstract: A reader for reading magneto-optical characters, with the options of printing them or storing them.A track is provided for driving a check from an inlet plate (T1) over various portions of path (T2, T3 and T4). Wheels (G1 and G2) are provided beneath the path, and may be motor driven. Above said wheels there are heads (TM1 and TM2) respectively for premagnetizating and reading CMC7 characters. Each head is mounted on one arm of a corresponding bell crank (L1, L2) which cranks are pivoted to a frame at their corners (L10 and L20). The other arm (L12 and L22) of said bell cranks are resiliently urged by springs (L15 and L25) mounted on an actuator rod (AC) which is itself driven by an electromagnet (EM2). It has turned out that such an assembly ensures effective transducer operation of the heads (TM1 and TM2) for reading CMC7 characters.

Abstract: A subsystem component for use in an image processing system to compute a gray scale histogram function or various statistical functions relating to the coordinates of a region or regions in a binary image. A selected function is computed at the video rate of frame generation.

Abstract: The invention relates to a process for fault detection on structures on surfaces of electrical components or the auxiliary means required to manufacture these components, wherein the actual structure provided in each respective case is scanned with a test beam. In accordance with the invention, provision is made for structure faults or structure deviations to be directly ascertained from sequences of picture points within the scope of an ordinate scanning.

Abstract: An image processing technique for detecting and measuring the size of bruises on fruit, such as apples, is disclosed. A line scan camera is employed to obtain a gray level image of each piece of fruit as it is rotated in front of the camera. The image is digitized, and transferred to a computer which performs a plurality of steps on the image to determine the sizes of any bruises on the fruit from the gray level data. The image data is filtered to remove interference due to pixel-to-pixel variations in the camera, background noise, etc., and is then thresholded to obtain a binary image representation. Clusters of zeros in the resulting binary image that are representative of potential bruises are analyzed to determine their size and shape. If the shape of the cluster is nearly circular, it is determined to be a bruise. From this information, the amount of bruise area on the fruit may be determined, and the fruit graded accordingly.

Abstract: In many digital image processing methods, it is desirable to selectively apply digital image processing to identifiable structures in the image. For this purpose it is known to select gray level thresholds between structures based upon the location of corresponding peaks in the gray level histogram of the digital image. However, it is a problem to automatically detect the locations of peaks in the histogram and to select the gray level thresholds. The present invention provides a method for automatically detecting the peaks and selecting gray level thresholds for segmenting a digital image into distinguishable structures including the steps of detecting peaks in a gray level histogram of the digital image by applying smoothing and differencing operators to the gray level histogram to generate a peak detection function wherein positive to negative zero crossings of the peak detection function represent the start of a peak, and maxima following such a zero crossing represents the end of a peak.

Abstract: Patterns (f) in a scene (S) are analyzed in order to automatically distinguish symmetrical perceptible areas (P) in the pattern (f), even if said symmetrical perceptible areas (P) are intersected by auxiliary patterns (F). A digitized image (Id) is used, and a test point (cO) located inside the digitized pattern (fd) is used as a starting point. A sequence of circles of increasing diameter is drawn around the test point. Whenever one such circle (CO.sup.i) intersects the digitized pattern (fd) a new center (c1) is chosen, offset from the preceding center (c0) in a direction going away from the point of intersection. By iteration, a sequence of test points (cn) is thus constructed which, for a symmetrical pattern, converges on its center of symmetry (c).

Abstract: A handwritten pattern approximated to a series of polygonal lines consisting of segments is compared with a candidate pattern selected from dictionary patterns stored in the memory, basing on the angular variation between adjacent segments of both patterns. If the difference between angular variations of adjoining segments of both patterns is outside of a certain range, it is tested whether the difference between an angular variation across three or more consecutive segments and the above reference angular variation is within the range.

Abstract: A pattern recognition apparatus includes a pattern preprocessing part having a plurality of outputs, and a clock oscillator, an information consolidator and a memory having a plurality of nonlinear oscillating circuits, respectively. Each of the oscillating circuits includes a plurality of inputs for controlling the oscillation condition and at least one output. In the information consolidator, the output of each oscillating circuit is interconnected to predetermined inputs of other oscillating circuits in a predetermined relation, and the inputs of the oscillating circuits are coupled to the outputs of the pattern preprocessing part in a predetermined distribution. The clock oscillator receives, as a main exciting signal, the sum of a group of signals obtained by phase-adjusting the output of a substantial part of oscillating circuits contained in the information consolidator.