Abstract: A method is disclosed for determining skeletons of any two dimensional images and three-dimensional skeletons of vessel trees. In the three dimensional case, the method takes a binary volume of any vessel tree as input and produces centred, connected and one-voxel wide skeleton of the input vessel tree. In the two dimensional case, centred, connected and one-pixel wide skeletons of input binary images are produced. The method includes the steps of: (1) calculating a 3D/2D distance transform, (2) locating local maximum voxels/pixels, one-voxels/pixel, (3) propagating skeletal elements to get a full set of skeletal elements, and (4) removing redundant local maximum voxels/pixels to obtain a one-voxel/pixel wide skeleton.

Abstract: In a method for automatic removal of image errors in digital image data, for example which have arisen due to scratches, hairs, etc., a contour mask is generated with a contour filter and a color mask is additionally generated that covers the image areas having the typical color of an image error. By operating the contour mask and the color mask, an error mask arises that is also automatically corrected and is manually edited as warranted. The remaining contours of the error mask are vectorized. An automatic bleeding retouch is then implemented along the vectorized contours.

Abstract: A novel wafer or circuit board automatic scanning optical inspection system and technique using the reference comparison principle for finding defects in scanned sample images, that enables a new degree of universality in finding all defects including very small single pixel conductor breaks and shorts and defects of irregular shapes, and together with simultaneous design rule processing, through the use of processed skeletal reference images and with separately programmable alignment tolerance and detection parameters.

Abstract: An adaptive multiscale method approximates shapes with continuous or uniformly and densely sampled contours, with the purpose of sparsely and nonuniformly discretizing the boundaries of shapes at any prescribed resolution, while at the same time retaining the salient shape features at that resolution. In another aspect, a fundamental geometric filtering scheme using the Constrained Delaunay Triangulation (CDT) of polygonized shapes creates an efficient parsing of shapes into components that have semantic significance dependent only on the shapes' structure and not on their representations per se. A shape skeletonization process generalizes to sparsely discretized shapes, with the additional benefit of prunability to filter out irrelevant and morphologically insignificant features.

Abstract: A method and apparatus for performing thinning, or skeletonization, of binary images or figures are provided. More particularly, a method is implemented in a system for encoding a figure using a mid-crack boundary representation to identify the boundary contour of the figure. Then, the boundary contour of the figure is traversed while the boundary is selectively displaced inward. The boundary contour is also concurrently re-encoded based on the selectively displaced boundary.

Abstract: A method and apparatus for overloading operators in Page Description Language (PDL) defined graphics redefines the graphic, such as to compensate for certain undesirable line growth characteristics of a target printer or display device. Line width can be controlled based on various attributes, such as color, size, or width. Additionally, the method and apparatus determine whether a line of an object of the graphic intersects with or completely overlays another object. If it does, only those overlying portions are redefined. A line of the graphic can be redefined to be wider, narrower or of a differing color to achieve a desired printed graphic. Also, the method and apparatus can compensate for fill in of reverse-line and “cookie-cutter” holes. The method, performed within a decomposer, examines individual lines of a PDL defined graphic, determines if compensation is needed to achieve a desired printed output, and redefines an operator to achieve the compensation.

Abstract: Methods and computer executable instructions are provided for converting original images such as photographic pictures into various other formats such as drawings by reducing the amount of data contained in the original images. It is a feature of this invention that a photographic picture can be converted into a drawing useful as a template to enable a person to paint the template into a likeness of the photographic picture. In a preferred embodiment, an original image is supplied to a computing environment where it is quantized into pixels of information. Each scalar value of each pixel is evaluated and aberrant pixels are absorbed into adjacent pixels by converting various scalar values. Thereafter, pixels having similar scalar values are grouped into regions. The boundaries between the regions are absorbed into each of the surrounding regions to provide a smoothed boundary. The regions then provide a template conducive to painting.

Abstract: A method and device for generating, displaying and manipulating three-dimensional images for medical applications is provided. The method creates a three-dimensional images from MRI or other similar medical imaging equipment. The medical imaging system allows a user to view the three-dimensional model at arbitrary angles, vary the light or color of different elements, and to remove confusing elements or to select particular organs for close viewing. Selection or removal of organs is accomplished using fuzzy connectivity methods to select the organ based on morphological parameters.

Abstract: The present invention relates to an outline font. The purpose lies in supplying a transforming and rendering method in the shape where the important form of the shape, such as the thickness of the stroke, will be maintained from the basic shape, and at the same time the whole size of the shape can be fitted into a given bounding box. The present invention supplies the transformation method of a handwriting font including searching for the medial axis of a given element, obtaining contact points by using maximal inscribed circles with the medial axis and points of the medial axis as a center, and applying the information about the medial axis points and the contact points in case of a transformation of the element in the transformation method of handwriting where a given element can be optionally transformed in sizes in order to achieve the purpose. The present invention enables the transformation of the shape.

Abstract: A digital image (J.sub.o) processing method for automatic extraction of strip-shaped objects, includes a skeletonization operation with steps for forming smoothed images (J.sub.i) at several scales (.sigma..sub.i), and, in each smoothed image (J.sub.i), extracting boundaries of objects, extracting potential median pixels (.OMEGA..sub.iP) associated with the center (.OMEGA..sub.i) of a circle of radius (R.sub.i) proportional (k) to the scale (.sigma..sub.i), tangent to boundaries at a pair of distinct pixels (E.sub.1, E.sub.2), and associated with a measure of dependability regarding alignment of the center (.OMEGA..sub.i) and pixels of the pair, extracting median pixels .OMEGA..sub.iM, and constructing skeletons of objects by tracking in the digital image (J.sub.MED) formed by the extracted median pixels. The step of extracting median pixels .OMEGA..sub.

Abstract: A facsimile apparatus contains transfer modes such as a standard mode, a fine image quality mode, a super fine image quality mode which are different from one another in resolution. A signal representing each transfer mode is transmitted from the transmission side to the reception side before the transmission and reception of facsimile image data. A control unit detects the resolution of reception data according to the transfer mode signal to switch over methods for conversion into high-tone in a high-tone conversion unit or to set the magnification of conversion of a resolution conversion unit according to the detected resolution. In this case, when high resolution is detected, the unit for conversion into high-tone executes conversion into multi-value using an area of 17.times.17 picture elements around candidate picture elements for processing. When low resolution is detected, the unit executes the conversion into multi-value using an area of 11.times.

Abstract: The present invention relates to a process for producing a skeleton from a binary image by using the compressed run data of the image instead of using natural expanded binary format of the image. A run length compressed image is a binary image (black/white) that is compressed by taking stripes of pixels of the same value and grouping them together into a single "run" of pixels. This compression technique saves significant amounts of memory on most binary images. It is an object of this present invention to enable the image to be skeletonized in its run length compressed format. This reduces the amount of data that needs to be stored in memory. Further, since the image is in "runs" and not pixels, the pixels that are not affected are not evaluated. By increasing available memory while reducing the amount of pixels to be evaluated, the speed for the process is increased on the order of several magnitudes.

Abstract: The document display system comprises a first character shape generator, a first display and a first transmitter in a first terminal device; and a first receiver, a second character shape generator and a second display in a second terminal device. The first character shape generator generates the character shapes of characters constituting at least part of a document. The character shapes are generated in response to skeleton information and in response to shape generation information that includes information representing a typeface. The first display displays the characters with the character shapes generated by the first character shape generator so that the at least part of the document is displayed set in the typeface. The first transmitter transmits character information indicating the skeleton information for each of the characters. The first receiver receives skeleton information for each of the characters.

Abstract: A method of generating an output image having a lower resolution than that of a source image is provided. A thinned image is generated by discarding outer pixels of the source image. In parallel with generation of the thinned image, an averaged image is generated, such that each pixel of the averaged image represents an average of a subset of pixels in the source image. A first output image is then generated by subsampling the pixels of the thinned image. A final output image is generated as a function of both the first output image and the averaged image.

Abstract: A method (30) is provided for the detection and analysis of anomalies (32) in a road surface (36). An image (34) of the road surface (36) is obtained (82) wherein traffic control markings (76) are masked (88). The image (34) is filtered (90) and a pixel map (92) is produced (98). The pixel map (92) is partitioned (112) into a multiplicity of subimages (108). For each subimage (108), anomaly parameters are identified (120) and a status characteristic is determined (124) and assigned (136). A subimage map (138) is produced (142) depicting the subimages (108) and their status characteristics. A determination (156) is made as to which subimages (108) contain anomalies (32). Anomaly-containing subimages (108) are grouped (158) into anomalous objects (152). For each anomalous object (152), an object type (162) is determined (160) and assigned (164). An object map (154) is then produced (166) depicting the anomalous objects (152).

Abstract: An image containing characters and lines, which is inputted by a scanner or the like, is binarized and stored in a memory. A thickening process is repeatedly applied on groups of significant bits stored in the memory, until a predetermined condition is fulfilled. After this process, the thickened image is thinned at least a greater number of times than the number of times the thickening process is repeated in order to eliminate lines. As a result, characters are determined to exist at areas where groups composed of remaining significant dots exist, then characters are separated from the input image based on the areas of groups composed of the remaining significant dots.

Abstract: A design for a high reliability recognition system utilizes two optimized thresholds for each class k of a prototype data base. One threshold is a class region threshold CR.sub.k and the other is a dis-ambiguity threshold DA.sub.k. CR.sub.k specifies a constrained region belonging to a class k, and DA.sub.k corresponds to a value with which a sample belonging to class k can be correctly recognized with a high level of confidence. During recognition, if the distance D(x, r.sub.M) between an input sample x and the representative prototype r.sub.M of a nearest class M is larger than the class region threshold CR.sub.M, x will be rejected. Furthermore, if the distance D(x, r.sub.M) is subtracted from the distance D(x, r.sub.S) between x and the representative prototype r.sub.S of a second nearest class S, the resulting distance difference must be greater than the dis-ambiguity threshold DA.sub.M, or x will be rejected. An inventive algorithm is used to compute optimum thresholds CR.sub.k and DA.sub.

Abstract: An apparatus and method for determining the Fiber Orientation Distribution (FOD) in an injection molded polymer is described. The transparent sample (12) is comprised of a matrix material, reinforcing fibers and opaque tracer fibers (12A) which are analyzed to determine the FOD and the fiber length of all fibers in the sample. The sample is formed in an injection molding apparatus (14) which includes an injection mold (16) and an injection molding unit (34). The sample is constructed and placed in the imaging apparatus (58) in order to be optically sectioned. The imaging apparatus includes an inverted microscope (60), a camera with a controller (62), a motion controller board (64) and a frame grabber (66). The sample is scanned along three (3) axes in x-y plane in the field of view of the microscope at increasing depths in the sample along the z-axis which is the line of sight of the microscope. The 3D images are analyzed to determine the FOD and fiber length of the fibers in the sample.

Abstract: Image processing apparatus is disclosed for splitting, for subsequent storage or processing by OCR apparatus, character images in digital form comprising connected characters, the image processing apparatus extracts an image skeleton from an input image and represents the topology of the image skeleton in a first data structure; generates, from the first data structure, second data structures representing the topologies of the results of possible splits of the input image skeleton; evaluates the second data structures and selects one of the possible splits to thereby split the input image.

Abstract: The present invention enables application of the bold process to thicken the data or the outline process to hollow the inside of the data to leave the contour thereof to the gray scale data representing the gray scale font data or image data. Using the gray scale modified data as a result of modification process can make sentences visually appealing. The gray scale modified data comprises an area judging means, a background color area processing means, and a non-background color area processing means. The area judging means judges whether each pixel on the gray scale data to be processed has the background color level. The background color area processing means sets a value calculated by the values of the neighboring pixels of each pixel on the background color area to the value of the mapping pixel on the modified data.

Abstract: A method for generating a vector representation of the skeletal configuration of an elongate graphic image from a raster representation thereof, the method comprising providing a gray scale raster representation of the graphic image, for at least one of a plurality of lines of the raster representation, performing the following processes at least once: employing the gray scale raster representation to determine two path endpoints and computing a midpoint between the two endpoints.

Abstract: Pattern recognition, for instance optical character recognition, is achieved by forming a skeletal representation (400) of a pattern (300), processing the skeletal diagram (400) to improve representation of curved lines (308) in the pattern (300), representing the processed skeletal diagram (500) by a connectivity matrix (602), and finding a minimum spectral distance between the connectivity matrix (602) and a set of template matrices corresponding to known patterns.

Abstract: The present invention has an object to provide an image processing method for generating an image through thinning from which the topological characteristics of the image can be extracted more precisely and more accurately by deleting vestigial lines, that is, breaks in a line.

Abstract: The techniques for determining the orientation of lines in images disclosed herein take as their starting point an apparently universal typesetting convention: that symbols are generally closer to other symbols in the line to which they belong than they are to symbols in other lines. The techniques make a graph by connecting symbols in an image of a block of text with edges (lines) and then determine the predominant orientation of the shortest edges. Because the symbols in a line are closer to each other than to symbols in other lines, the predominant orientation of the shortest edges will be the same as the orientation of the lines. An advantageous mode of obtaining the shortest edges is to construct a Euclidean minimum spanning tree from the graph. A Euclidean minimum spanning tree of a graph contains all the nodes (symbols) of the original graph but connects the nodes with the shortest possible set of edges. The edges which predominate in the Euclidean minimum spanning tree are the shortest edges.

Abstract: An image processing method for thinning lines converts pixels having a density of a configuration to a background density. Pixels are converted according to conditions which include densities of pixels in a neighborhood around a pixel to be processed. Pixels to be converted to the background density are first converted to an index value. Subsequent pixels are evaluated using the index values of adjacent, previously evaluated pixels. Index values are then converted to the background density.

Abstract: A character bit map is skeletonized using competitive locally independent processes (CLIPs) by defining for each foreground pixel an individual CLIP with a center location and radius defining the pixels included in the CLIP. In iterative cycles, each CLIP varies its center location and radius over predetermined ranges to produce a predetermined number of permutations thereof, computes a trial fitness value for each permutation, the trial fitness value being rewarded and penalized in accordance with foreground and background pixels included in the CLIP and penalized for each pixel included in both the CLIP and in another CLIP of greater fitness value. Each CLIP adopts the maximum trial fitness value as its next fitness value along with the corresponding center location and radius. Each CLIP is eliminated if its next fitness value falls below a predetermined minimum value.

Abstract: An image figure editing system capable of manipulating figure elements given in terms of the image data in a substantially similar manner as the editing of the figure elements given in terms of the vector data. In the system, the position data to specify the figure elements to be operated in the figure are inputted, one of operation modes is specified, characteristics indicative of a shape of the figure are extracted, and features of the figure are determined according to the extracted characteristics, by carrying out an operation indicated by the specified operation mode to the figure elements specified by the inputted position data. In addition, an image of the figure is modified according to the determined features and the extracted characteristics, by carrying out the operation indicated by the specified operation mode to the figure elements specified by the inputted position data.

Abstract: A system for recognizing handwritten words of cursive script by extracting temporal information from static text is described. The temporal information consists of the shape, location and temporal sequence of the segments, or "primitives", constituting the word and is translated from bit maps images of the words into digital information format. By comparing the digital information with formatted digital information for a plurality of known words, as may be stored in computer memory, a list of words having a relatively high probability of matching the static text is generated, and used for word recognition by a process of selecting the most likely match.

Abstract: A dividing step (a) divides into segments at least one character line forming input characters to be read. A partial-figure forming step (b) forms partial figures by combining segments from among the thus obtained segments. A character reading step (c) attempts to read each of said partial figures as a character. A network forming step (d) forms a network wherein partial figures, among the partial figures, which have been read in said character-reading step (c) are used as nodes and said nodes are connected with one another by links, and wherein said links respectively have appropriate weights. An optimum-path selecting step (e) selects an optimum path from among paths existing in said network so that said optimum path comprises nodes, from among said nodes, which respectively correspond to said input characters.