Abstract: A method and apparatus for automatically determining pathway boundaries for autonomous robotic navigation. A structured light pattern consisting of a plurality of lines is projected on the pathway and background and the reflected pattern is detected by a stereoscopically located video camera. Since the texture of the pathway and background are different, the captured image may be processed to distinguish one from the other. Image processing includes Hough transformation of the image field, rank filtering and morphological dilation. The end points of the pathway lines identified by such processing as well as the 3-D coordinates of such end points are determined.

Abstract: Tone values of a photographic image are stored in a frame store and retrieved from the frame store to produce a corresponding video display. A mouse or trackball is used to locate a point on display and a gamma correction curve look up table is used to interactively show the effect of color correction. Adjustment can be made via lighter and darker switches on the mouse or trackball. These switches subtract or add a bump function to the gamma correction curve in the selected color density plane. The bump function is displaced by the measured tone value in order to achieve more flexible control over the color correction.

Abstract: An image processing apparatus comprises: first binarizing means for binarizing input image data by an average value based on the binarized data of pixels near an objective pixel which has already been binarized; second binarizing means for binarizing the input image data by a predetermined threshold vlaue; and selecting means for selecting either one of the first and second binarizing means in accordance with whether the input image data indicates an edge portion or not, whereby by selecting either a binarizing process by the average density preservation method or a binarizing process by a fixed threshold value in accordance with the presence or absence of an edge of input data, an image of a high grade can be reproduced for either a half tone image or a character image.

Abstract: A character recognition device has a subdivider, a features calculator and a character code recognition device. Image data for a single character area is extracted from scanned character image data and input to the subdivider. This subdivider divides the image data for the single character area into subregions. The features calculator calculates quantified features in each subregion based on a degree of resemblance between a template and image data in the subregions. When the features of each subregion are calculated for all subregions constituting the single character area, a character code corresponding to the scanned character image data is recognized by the character code recognition device based on the quantified features of each of all subregions.

Abstract: An image data filing system capable of facilitating an easy and accurate modification on the image filed in the system, without demanding a cumbersome operation for an operator. In the system, the image is modified by specifying an image to be modified among images filed in the system; displaying the specified image to be modified; specifying a region to be modified on the displayed image to be modified; obtaining a hard copy of the specified region to be modified; making a modification on the obtained hard copy of the specified region to be modified to obtain an modified image; and entering the modified image obtained on the hard copy into the system.

Abstract: An image coding apparatus and method are disclosed. The method and apparatus are applicable to coding any array of arbitrary dimension. The apparatus utilizes FIR filters which generate three or more sets of coefficients from linear arrays generated from the array being coded. Each set of coefficients represents spatial frequency information of different frequencies. The filters utilize coefficient sets which are preferrably orthogonal.

Abstract: The present invention relates to a primary color forming circuit for a complementary color CCD in which complmentary color pixels are arrayed in a predetermined pattern to form lines which have the same pattern as every other line. The output of one line is delayed by one line drive period and combined with the output of the next line to form source signals. Pseudo primary colors are obtained by combining these source signals with appropriate coefficients. Therefore, the present invention provides an image pickup apparatus which has advantages of both a complementary color filter CCD and a primary color CCD.

Abstract: Computer programs for capture and analysis of digitized image data signals comprising frame, scan line and digitizer stripe numbers, and edge-detection or bright-spot data relating to images in a field of interest, which includes the steps of converting the edge-detection data to bright spot, reducing the amount of data processed by eliminating duplicate data, normalizing the data into X and Y coordinates, finding image features of interest, selecting features to be reported, and reporting type, number, pattern or location of the image features in the field. An example is processing of shotgun target images scanned by a video camera (whose signals are digitized) through a slit viewing a rotating illuminated drum on which a target has been placed. Other uses include: astronomical image analysis; satellite photos of sites; gatherings of people, things, animals or crops in fields; and movements of objects through a field of view such as objects on a conveyor belt, traffic, ballistic objects, and the like.

Abstract: Line segments in stored picture data derived from an optically scanned document are detected by detecting line elements in the picture data, determining if the length of a line element is greater than a threshold length l.sub.th, determining a connecting relationship of one line element of greater than threshold length to another line element of greater than threshold length, and identifying a line segment in accordance with the connecting relationship of the line elements. Predetermined patterns formed of the detected line segments are detected by sensing if two line segments adjoin each other, determining the adjoining relationship of those two line segments, and identifying the pattern which includes those two line segments as a function of their adjoining relationship.

Abstract: An original-reading device includes a light source for irradiating an original disposed at a reading position, a photoelectric conversion device for converting light reflected from the original into an image signal, a reading unit supporting the light source and the photoelectric conversion device, a feeder for feeding the original to the reading position, a reference member provided at a position different from the reading position, for correcting an image signal from the original, and means for moving the reading unit between a position for reading the reference member and the reading position.

Abstract: A matching machine is provided for use in model-based entity recognition systems. Input to the matching machine includes a detected data event separated into portions and a set of predefined hierarchical models of desired entities. The matching machine employs recursive procedures to match data event portions against model parts, and records in list structures matches between event portions and model parts. For each match, a degree of match is calculated according to a function defining the involved model part. An essential feature of calculating a degree of match is the accumulation of working degrees of match (i.e. figures of merit reflecting the strength of match of a subset of data event portions to a subset of model parts and the degree to which the relationship of portions matches the predefined relationship of model parts in arbitrarily complex models). The degree of match is recorded in the list structures.

Abstract: In apparatus for successively inspecting an article, for example a telephone cable connector having a plurality of two dimensionally arranged connecting pins, the article is successively conveyed to and mounted on an inspecting table, an image pick-up camera is stepwisely moved for photographing all portions of the connector, and the direction of movement of the camera is alternately reversed.

Abstract: A pattern recognition apparatus is disclosed that is capable of high-speed, automatic recognition of input image patterns with a high degree of accuracy and which also has a broad range of operational tolerance to shifts in the position of input image patterns. The pixel data of an input image are compared by separate circuits with pixel data of reference image patterns. Other circuits use these comparisons to select that one of the reference patterns to which the input image corresponds.

Abstract: A feature extracting method converts an image data of an input line drawing into a graph representation and recognizes the input line drawing by use of the graph representation. The feature extracting method comprises a first step of equally dividing the image data of the input line drawing into meshes, a second step of obtaining a cell by integrating one or more meshes, and a third step of recognizing a cell pattern for each of cells which are obtained so as to form a graph in correspondence with each of the cells.

Abstract: A pattern inspection technique and apparatus, suitable for wafer and printed circuit board and related applications, employing novel intelligent imaged-pattern shrinking and expanding architecture to identify permissible line widths, spacing and in surrounding material context, and to identify defects or errors.

Abstract: An image processing system for extracting a plurality of local images each composed of m.times.n pixels, successively from a two-dimensional original image, to process the extracted local images and to convert the original imae into a second image composed of modified pixel data corresponding to the local images is disclosed which system includes a plurality of local image extracting circuits for temporarily storing time-sequentially applied pixel data in each local image extracting circuit while shifting the storage position thereof, to deliver m.times.

Abstract: An image processing apparatus switchable between an observation mode for projecting an image onto an observation position and a read mode for projecting the image onto a read position includes an optical system for forming an optical path for projecting the image onto the observation position. The optical system includes at least one movable mirror for receiving a light beam from the image. The mirror is movable to a first position for reflecting the light beam from the image toward the observation position and to a second position for reflecting the light beam toward the read position.The apparatus also includes a first moving mechanism for moving the mirror to the first position and to the second position, an image sensor disposed at the read position and movable so as to enter into the optical path or to retreat from the optical path, and a second moving mechanism for moving the image sensor along a straight line crossing the optical path.

Abstract: A sequence of documents is delivered to an optical scanner in which each document is scanned to form a digital image representation of the content of the document. In one embodiment, the image representation is converted into code (ASCII) and is automatically examined by data processing apparatus to select search words which meet predetermined criteria and by which the document can subsequently located. In another embodiment, the image is not converted. The search words are stored in a nonvolatile memory in code form and the entire document content is stored in mass storage, either in code or image form. Techniques for selecting the search words are disclosed.

Abstract: Apparatus for generating an array of control data, the array having a number of pixels each of which corresponds to a respective pixel in an image, the apparatus comprising a first store (100) for storing the image data; a second store (17) for storing the control data; and a processor which, for each control data array pixel generates and stores in the second store (17) a control value in accordance with a predetermined algorithm which defines the control value as a function of the corresponding image pixel color components and the corresponding color component(s) of the target colors, wherein the generated control values can be ordered on a scale of at least three value classes.

Abstract: A method storing multiple scanned images of a response-marked form document comprises scanning a copy of the form document on which no response information has been marked, using a scanning method that detects substantially all visual information present on the form document and produces a pixel image thereof and storing that pixel image data in a computer storage device. Additonal steps comprise scanning at least one copy of the same form document that has been marked with at least one response mark that absorbs light falling outside the visible spectrum, using a scanning method that detects only the presence of light outside the visible spectrum and that produces a pixel image of said at least one response mark, then compressing that pixel image to reduce the amount of pixel data used to repressent areas of the form document bearing no response marks and storing the compressed pixel image data in a computer storage device.