Local Neighborhood Operations (e.g., 3x3 Kernel, Window, Or Matrix Operator) Patents (Class 382/205)
-
Patent number: 6600839Abstract: A filter reduces artifacts, such as grid noise and staircase noise, in block-coded digital images with image block boundaries. The type of filtering is determined after an estimation of the image global metrics and local metrics. For areas of the image near grid noise, the filter performs low pass filtering. For image fine details, such as edges and texture, no filtering is performed so that masking is avoided. The filter operates in intra-field mode and uses a fuzzy logic process, pixel deltas, and dual ramp generators to determine the horizontal and vertical length of a processing window surrounding an image block boundary.Type: GrantFiled: May 17, 2001Date of Patent: July 29, 2003Assignee: STMicroelectronics, Inc.Inventors: Massino Mancuso, Antonio Maria Bornco
-
Patent number: 6597813Abstract: A technique combines most of the advantages of both blue noise and clustering in generating a partially clustered aperiodic mask and using it in a dither array algorithm for halftoning, such that the amount of clustering can be modulated. A method to create an enlarged partially clustered aperiodic dither mask, appearing as a modulated version of a periodic enlargement of original mask, is used when the original mask generates patterns that are not visually pleasing or the original mask doesn't generate enough distinct grey levels. The method of halftoning of grey scale images utilizes a pixel-by-pixel comparison of the image against a partially clustered aperiodic mask in which the clustered aperiodic mask is comprised of a partly random and partly deterministic single valued function which is designed to produce visually pleasing dot configurations when thresholded at any level of grey.Type: GrantFiled: March 11, 1999Date of Patent: July 22, 2003Assignee: International Business Machines CorporationInventors: Mikel J. Stanich, Gerhard R. Thompson, Charles P. Tresser, Chai W. Wu
-
Publication number: 20030123094Abstract: Methods and apparatus are provided to smooth a jagged outline of a reconstructed text generated through any of a variety of systems, such as, for example, an image scanner. Given the reconstructed text with a jagged outline, the methods smooth the outline of the text first through location of horizontal and vertical edges, and subsequent gray scale reconstruction of the edge ink level from a one-bit or a low number of bits edge pattern. The methods then provide identification of the local text outline pattern and corresponding ink level adjustment of the center pixel. The methods produce text having very smooth horizontal and vertical edges. The methods then smooth slanted edges, mend holes in the text, and remove isolated pixels of ink.Type: ApplicationFiled: December 11, 2002Publication date: July 3, 2003Inventors: Ron J. Karidi, Lai Chee Man
-
Patent number: 6587586Abstract: A method for extracting an image representing textual information from a video sequence includes the following steps. First, receiving a sequence of video frames, each including an image of textual information. Then, locating the textual information in each frame of the video sequence to form a stack of text arrays, each array containing data representing substantially only the textual information. Finally, extracting a single textual image array representing the image of the textual information from the stack of text arrays. Apparatus for extracting an image representing textual information from a video sequence includes a source of a video sequence having a plurality of frames, each containing an image of the textual information; and a processor, coupled to the video sequence source, responsive to all of the plurality of frames, for generating a single array representing an image of the textual information.Type: GrantFiled: June 12, 1997Date of Patent: July 1, 2003Assignee: Siemens Corporate Research, Inc.Inventors: Yuntao Cui, Qian Huang
-
Patent number: 6577744Abstract: A robust means of watermarking a digitized image with a highly random sequence of pixel brightness multipliers is presented. The random sequence is formed from ′robust- watermarking-parameters′ selected and known only by the marker and/or the marking entity. A watermarking plane is generated which has an element array with one-to-one element correspondence to the pixels of the digitized image being marked. Each element of the watermarking plane is assigned a random value dependent upon a robust random sequence and a specified brightness modulation strength. The so generated watermarking plane is imparted onto the digitized image by multiplying the brightness value or values of each pixel by its corresponding element value in the watermarking plane. The resulting modified brightness values impart the random and relatively invisible watermark onto the digitized image. Brightness alteration is the essence of watermark imparting.Type: GrantFiled: June 16, 1998Date of Patent: June 10, 2003Assignee: International Business Machines CorporationInventors: Gordon Wesley Braudaway, Frederick Cole Mintzer
-
Patent number: 6574367Abstract: An apparatus for pattern matching encoding includes a pattern extracting unit for extracting patterns in input binary or multi-level image data, an accumulating/match-checking unit for accumulating each extracted pattern and executing a match-check thereof with patterns that have been accumulated, a line segment checking unit for checking each pixel in the match pattern or the extracted pattern as to which line segment in directions the pixel is positioned on, and an encoding unit for encoding the extracted pattern by selecting a template having an optimum form according to the result of the line segment checking. The codes in pattern matching encoding are reduced, and the efficiency of code transmission and code accumulation is improved.Type: GrantFiled: January 7, 1998Date of Patent: June 3, 2003Assignee: NEC CorporationInventors: Mitsutoshi Arai, Takahiro Hongu, Kouichirou Hirao
-
Patent number: 6574374Abstract: A method including defining a morphological filtering operator designed to remove ringing artifacts in a decompressed image, and then applying that operator upon pixels of the decompressed image that belong to non-texture image regions.Type: GrantFiled: April 14, 1999Date of Patent: June 3, 2003Assignee: Intel CorporationInventor: Tinku Acharya
-
Patent number: 6570616Abstract: When processing an image of a two-dimensional pixel arrangement such as a Bayer arrangement, when a pixel G (j, i) of a first color component which represents luminance is interpolated, the value of the first color component associated with the primary reference pixels located adjacent to the top and bottom or left and right of the interpolated pixel position, and secondary reference pixels of the first color component, located in the vicinity of the primary reference pixel, is checked. The existence of an image edge component is detected. When an image edge component is detected, the weight of the reference pixels, for example, G(j−1, i), G(j+1, i), G(j, i−1) and G(j, i+1) is changed at the top, bottom, left and right.Type: GrantFiled: October 16, 1998Date of Patent: May 27, 2003Assignee: Nikon CorporationInventor: Zhe-Hong Chen
-
Patent number: 6549646Abstract: A divide-and-conquer (DAC) method and system improve the detection of abnormalities, like lung nodules, in radiological images via the use of zone-based digital image processing and artificial neural networks. The DAC method and system divide the lung zone into different zones in order to enhance the efficiency in detection. Different image enhancement techniques are used for each different zone to enhance nodule images, as are different zone-specific techniques for selecting suspected abnormalities, extracting image features corresponding to selected abnormalities, and classifying the abnormalities as either true or false abnormalities.Type: GrantFiled: February 15, 2000Date of Patent: April 15, 2003Assignee: Deus Technologies, LLCInventors: Hwa-Young Michael Yeh, Jyh-Shyan Lin, Yuan-Ming Fleming Lure, Xin-Wei Xu, Ruiping Li, Rong Feng Zhuang
-
Publication number: 20030067976Abstract: An apparatus has a one-dimensional DCT transformer (101) for applying the one-dimensional orthogonal transforms to n inputs and outputting n coefficients, a transposition converter (102) for transposing n×n coefficients output from the one-dimensional DCT transformer (101) and outputting every n outputs, a multiplexer (103a) for selecting one output (1004a) of the n outputs from the transposition converter (102) or one output (1003a) of the n outputs from the one-dimensional DCT transformer (101), and a selector (100) for selecting data as a combination of one output (1010a) selected by the multiplexer (103a) and remaining (n−1) outputs, which are not input to the multiplexer (103a), of the n outputs from the transposition converter (102), or the input image data, and supplying the selected data as n data to the one-dimensional DCT transformer (101).Type: ApplicationFiled: October 4, 2002Publication date: April 10, 2003Inventor: Susumu Igarashi
-
Patent number: 6546134Abstract: A biometric machine method for instantaneous assessment of the fine motor control of a human individual through analysis of that individual's handwriting dynamics in which the dynamic data concerning forces, accelerations, and the like, of a scriber, is collected during the process of cursive writing. The criteria of stability, smoothness and synchronization of the writer's motion are returned by the system as quantifying measures of the neurological function. The method processes behavioral random signals with application of the correlation function analysis to the handwriting dynamic signals. In one embodiment, the system includes an instrumented pen connected to a computer, power supply, and analog-to-digital converter, and with the handwriting samples provided on any writing surface. In a second embodiment, the system comprises an instrumental digital tablet having an active area and non-instrumented scriber with the handwriting samples provided on the active area of the grid.Type: GrantFiled: March 28, 2000Date of Patent: April 8, 2003Inventors: Ruth Shrairman, Alexander Landau
-
Patent number: 6535632Abstract: Adaptive noise filtering is applied to an image frame of HSI data to reduce and more uniformly distribute noise while preserving image feature edges. An adaptive spatial filter includes a plurality of averaging kernels. An appropriate kernel is selected for each pixel for each of the hue and saturation components. A set of thresholds are defined for selecting the kernel for the hue component. Another set of thresholds are defined for selecting the kernel for the saturation component. The kernel for the saturation component is selected by comparing the intensity component to the saturation component thresholds. The kernel for the hue component is selected by comparing the product of intensity component and the saturation component to the hue component thresholds. A color gradient operation is applied to the filtered HSI data to aid in detecting image object boundaries. Object segmentation and other image processing techniques may be performed on the filtered HSI data.Type: GrantFiled: August 30, 1999Date of Patent: March 18, 2003Assignee: University of WashingtonInventors: Hyun Wook Park, Lakshmanan Gopishankar, Shijun Sun, Yongmin Kim
-
Publication number: 20030026495Abstract: Provided is a method and apparatus of processing an image using filters. The method and apparatus receives an input pixel and a pixel window associated with the input pixel from the image, classifies the input pixel using the pixel window into a range of classes identifying pixels suitable for various degrees of smoothing and sharpening operations, receives parameter independently set for sharpening and smoothing the image, and selects a filter for processing the input pixel based upon the pixel classification and the parameter settings for sharpness and smoothness.Type: ApplicationFiled: May 1, 2002Publication date: February 6, 2003Inventors: Jay Stephen Gondek, Amanda Jean Gillihan, C. Brian Atkins
-
Publication number: 20030012439Abstract: A method, apparatus, and article of manufacture provides a zero-crossing filtering region process for extracting data and related image elements from an input image. The input image is processed to generate connected component regions that correspond to image elements within the input image. The connected component regions are generated using a zero-crossing filter that attempts to identify pixels having a significant rate of change in value from their respective neighboring pixels. The connected components are then classified as being part of a foreground image or a background image that allows the image elements to be separated from the input image. Once the input image has been separated into multiple images, each separated image may be processed in a different manner depending upon the separate image's characteristics and uses.Type: ApplicationFiled: June 21, 2001Publication date: January 16, 2003Applicant: Microsoft CorporationInventor: Daryl Lawton
-
Publication number: 20030007187Abstract: The result of integrating a halftone dot-shape can be clustered. When the edge of the halftone dot-shape extends across a “vertical” edge of a window, the window is altered until the edge of the halftone dot-shape no longer crosses either vertical edge of the altered window. The image density of the portion of the halftone dot-shape contained within the altered window is determined. The block of image density of the portion of the halftone dot shape contained within the altered window is aligned with the right edge, the left edge, split between the left and right edges, or clustered in the “center” of the altered window. An amount of the determined block that extends into the original sample window is determined. This amount extending into the original window determines the amount of image density to be generated in the final output image based on that sample location of the window.Type: ApplicationFiled: June 29, 2001Publication date: January 9, 2003Inventor: Douglas N. Curry
-
Patent number: 6498868Abstract: A system for computing color information of a scaled version of a pixel of interest, includes a pixel identifier for identifying pixel color values for pixels in a neighborhood. The neighborhood includes the pixel of interest. The pixels in the neighborhood form a pattern. A transformer transforms the pixel color values of the pixel of interest and the plurality of neighboring pixels to produce a plurality of transformed values. A geometric pattern identifier identifies a geometric pattern which resembles the pattern formed by the pixels in the neighborhood. The geometric pattern is identified as a function of the transformed values. A selector selects an appropriate scaling algorithm to scale the geometric pattern. The algorithm produces the scaled version of the pixel of interest, along with the color information for the scaled pixel.Type: GrantFiled: June 11, 1998Date of Patent: December 24, 2002Assignee: Xerox CorporationInventor: R. Victor Klassen
-
Patent number: 6490375Abstract: An efficient and reliable method that identifies possible reference sites in an image is disclosed. The method emphasizes customizing the identification of reference sites for each application by providing for initialization of various parameters, such as the size of the reference site. The method then uses a series of measurable parameters to filter possible reference sites in the image and produces and ordered set of possible reference sites. The highest-ranking site of the ordered set is expected to be chosen as the reference site for production. A preferred implementation is disclosed that uses multi-resolution images to enhance efficiency of the identification and that specifically measures the symmetry, orthogonality and uniqueness of the windows.Type: GrantFiled: May 21, 2001Date of Patent: December 3, 2002Assignee: Cognex CorporationInventors: Arman Garakani, Leonid Taycher
-
Patent number: 6490374Abstract: The method identifies a Pixon element, which is a fundamental and indivisible unit of information, and a Pixon basis, which is the set of possible functions from which the Pixon elements are selected. The actual Pixon elements selected from this basis during the reconstruction process represents the smallest number of such units required to fit the data and representing the minimum number of parameters necessary to specify the image. The Pixon kernels can have arbitrary properties (e.g., shape size, and/or position) as needed to best fit the data.Type: GrantFiled: August 22, 2001Date of Patent: December 3, 2002Assignee: The Regents of the University of CaliforniaInventors: Richard Puetter, Amos Yahil
-
Patent number: 6483943Abstract: One or more contours of a picture element (PEL) array pattern are detected and, then, a strength of concavity/convexity on each contour (acuteness) and a direction of concavity/convexity are determined. Using an acuteness value of a contour point of interest, its curvature is classified into a group of strong curvatures or another group of weak curvatures such that the definitions of a direction of the concavity/convexity (contour direction) is appropriately switched from one to another for calculation. In a preferred aspect of this invention, using an acuteness value of a contour point of interest, its contour curvature is classified into either one of five curvature groups, comprising “strong concavity,” “weak concavity,” “straight line/inflection point,” “weak convexity,” and “strong convexity.Type: GrantFiled: September 30, 1999Date of Patent: November 19, 2002Assignee: International Business Machines CorporationInventor: Hiroyasu Takahashi
-
Patent number: 6473202Abstract: An image processing apparatus includes: a parameter computing means which computes the maximum density value, the summation of density differences, the average density and the maximum density difference for each pixel in a mask containing a target pixel; and a color determination counting circuit which makes appropriate, selective threshold comparisons based on the computed result from the parameter computing circuit to determine an area in a mask as being chromatic, achromatic or ‘the other’ area.Type: GrantFiled: May 19, 1999Date of Patent: October 29, 2002Assignee: Sharp Kabushiki KaishaInventors: Toshihiro Kanata, Yasushi Adachi, Makoto Otsu
-
Patent number: 6466695Abstract: The invention provides an apparatus, and related method, for providing a procedure to analyze images based on two-dimensional shape primitives. In the procedure, an object representation is created automatically from an image and then this representation is applied to another image for the purpose of object recognition. The features used for the representation are the two type of two-dimensional shape primitives: local line segments and vertices. Furthermore, the creation of object representations is extended to sequences of images, which is especially needed for complex scenes in which, for example, the object is presented in front of a structured background.Type: GrantFiled: August 4, 1999Date of Patent: October 15, 2002Assignee: Eyematic Interfaces, Inc.Inventors: Michael Pötzsch, Norbert Krüger, Christoph von der Malsburg
-
Patent number: 6453068Abstract: A method and apparatus for digital image processing are provided for controlling luminance channel overshoot in response to FIR or other enhancement processing based upon chrominance. The luminance value of the enhanced digital image data for each pixel of an image is compared to corresponding local minimum and maximum luminance values of the non-enhanced image obtained from the neighborhood of the subject pixel. If an enhanced luminance value for a subject pixel falls outside the corresponding local minimum/local maximum luminance value range by a positive or negative overshoot amount, the enhanced luminance value is adjusted to attenuate the overshoot amount. The amount by which the enhanced luminance value for an enhanced pixel is adjusted varies depending upon the chrominance of the enhanced pixel. Also, the amount of overshoot attenuation can vary depending upon whether the enhanced luminance value is greater than the local maximum luminance value or less than the local minimum luminance value.Type: GrantFiled: September 17, 1999Date of Patent: September 17, 2002Assignee: Xerox CorporationInventor: Xing Li
-
Patent number: 6445834Abstract: An image query and storage apparatus and method including a plurality of dynamically linkable feature modules is disclosed. Each of the plurality of feature modules extract a different set of feature information from an input image. The method and apparatus further includes a database coupled to the plurality of feature modules. The database includes storage for the different set of feature information for each of the plurality of feature modules. The method and apparatus support a query by image content of the database of images using the dynamically linked plurality of feature modules. The method and apparatus further includes a plurality of dynamically linkable scoring modules for processing feature specific scoring information generated by the feature modules.Type: GrantFiled: October 19, 1998Date of Patent: September 3, 2002Assignees: Sony Corporation, Sony Electronics Inc.Inventor: Hawley Knox Rising, III
-
Publication number: 20020118391Abstract: Disclosed are an image processing apparatus that can decrease such a misjudgment that an area including an extremely small-sized character is judged to be a halftone-dot area when a judgment of a halftone-dot area is performed based on a number of isolated pixels in a predetermined area, and an image forming apparatus that utilizes the image processing apparatus. An isolated pixel signal S2 that is outputted from *an isolated pixel detection unit 443 and that indicates whether a target pixel corresponds to an isolated pixel is inputted into a filter processing unit 444. The filter processing unit 444 corrects the isolated pixel signal S2 based on whether the target pixel should be handled as an isolated pixel. A judgment is then performed as to whether an area including the target pixel is a halftone-dot area, based on a signal S3 outputted from the filter processing unit 444.Type: ApplicationFiled: February 14, 2002Publication date: August 29, 2002Inventors: Taisuke Akahori, Hiroyuki Suzuki
-
Patent number: 6438265Abstract: A method of binarization used in an OCR system involves in determining text pixels by checking, for each pixel, that the difference between its value and the values of a plurality of pixels located at a predetermined distance therefrom is greater than a relative threshold corresponding to the difference in intensities between the text and the background of the image, subsampling the image at a rate corresponding to at least two pixels in order to detect kernels of text, and then binarizing the image pixels only in tiles of several stroke width sides containing text kernels by using in each tile, an absolute threshold estimated in that tile.Type: GrantFiled: May 12, 1999Date of Patent: August 20, 2002Assignee: International Business Machines Corp.Inventors: Andrei Heilper, Yaakov Navon, Eugene Walach
-
Patent number: 6430315Abstract: An image processing method which includes the acquisition (10) of an intensity image (J) formed by a matrix of points (A) having non-binary intensity values and representing filiform objects, and also includes steps for the detection of filiform objects, which steps include: automatic determination (20) of an intensity vector field ({right arrow over (V)}), forming a state image constituted by characteristics concerning magnitudes and angles [∥{right arrow over (V)}∥,&thgr;] of intensity vectors a the various points (A) of the intensity image, an operation (30) for chaining the points of an individual filiform object by means of automatic filtering [Q(i)] of the state image, thus performing a selection of a dense and continuous field of intensity vectors corresponding to points of the filiform object.Type: GrantFiled: May 20, 1998Date of Patent: August 6, 2002Assignee: Koninklijke Philips Electronics, N.V.Inventor: Shérif Makram-Ebeid
-
Patent number: 6429875Abstract: Image frames are stored as a plurality of pixel values. An image frame is viewed in which a perceived object (301) is to be distinguished from the remainder of the image frame. Input signals identify control positions in proximity to the boundary of said object. Pixel values are compared around each of the control positions to define a local color transition. Color transitions for the whole object are then produced by a process of interpolation. In this way, a process similar to chroma-keying may be performed without the need to record a foreground object against a specific type of background object. The chroma-keying process is performed locally with respect to a local color transition. This results in local control signals being produced and control signals for the whole object are then produced by a process of interpolation.Type: GrantFiled: April 1, 1999Date of Patent: August 6, 2002Assignee: Autodesk Canada Inc.Inventors: Daniel Pettigrew, Patrick Arbez
-
Patent number: 6427024Abstract: 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.Type: GrantFiled: April 2, 1999Date of Patent: July 30, 2002Assignee: Beltronics, Inc.Inventor: Robert Bishop
-
Publication number: 20020094129Abstract: A technique is provided that identifies screen and non-screen regions of a projected or displayed image to smooth and selectively remove moiré from the screen regions while maintaining sharpness in the non-screen regions. Each pixel in the image is classified as a screen or non-screen pixel and then pixels in a predetermined surrounding area of each pixel are examined to check the classification of that pixel. A low pass filter is applied to pixels in the image, such that, when the low pass filter is applied, one or more pixels covered by the low pass filter are respectively replaced by one or more other pixels covered by the low pass filter based on the examination.Type: ApplicationFiled: January 18, 2001Publication date: July 18, 2002Inventors: Jau-Yuen Chen, Joseph Shu
-
Publication number: 20020094128Abstract: A technique is provided that identifies screen and non-screen regions of a projected or displayed image to smooth and selectively remove moiré from the screen regions while maintaining sharpness in the non-screen regions. Each pixel in the image is classified as a screen or non-screen pixel and then pixels in a predetermined surrounding area of each pixel are examined to check the classification of that pixel. A low pass filter is applied to pixels in the image, such that, when the low pass filter is applied, a center of the low pass filter is selectively shifted relative to a current pixel based on the examination.Type: ApplicationFiled: January 18, 2001Publication date: July 18, 2002Inventors: Joseph Shu, Jau-Yuen Chen, Tsung-Nan Lin
-
Patent number: 6405203Abstract: A system, method, and article of manufacture is provided for tracking the distribution of content electronically. First, an electronic storage medium tracking identifier is incorporated onto an electronic storage medium and stored on a database. Next, a package tracking identifier is situated onto a package in which the electronic storage medium is stored. The electronic storage medium is then tracked while being shipped between various entities using the tracking identifier on the package. Further, the electronic storage medium may be identified using the tracking identifier on the electronic storage medium in order to afford various advertising, security, support, or retail-related features.Type: GrantFiled: April 21, 1999Date of Patent: June 11, 2002Assignee: Research Investment Network, Inc.Inventor: Todd R. Collart
-
Patent number: 6404908Abstract: A method an apparatus for detecting lines in medical images is disclosed, wherein a direction image array and a line image array are formed by filtering a digital image with a single-peaked filter, convolving the resultant array with second order difference operators oriented along the horizontal, vertical, and diagonal axes, and computing the direction image arrays and line image arrays as direct scalar functions of the results of the second order difference operations. Advantageously, line detection based on the use of four line operator functions along the horizontal, vertical, and diagonal directions in accordance with the preferred embodiments actually results in fewer computations than line detection based on the use of three line operator functions. In particular, because of the special symmetries involved, 3×3 second order difference operators may be effectively used.Type: GrantFiled: May 28, 1998Date of Patent: June 11, 2002Assignee: R2 Technology, Inc.Inventors: Alexander C. Schneider, Jimmy R. Roehrig
-
Patent number: 6377711Abstract: Trapping methods and systems for raster images along horizontal, vertical or diagonal directions uses edge detection for detecting edges between pixels and only diagonally neighboring pixels.Type: GrantFiled: June 30, 1999Date of Patent: April 23, 2002Assignee: Xerox CorporationInventor: Stephen C. Morgana
-
Patent number: 6369911Abstract: To convert gray-scale data of each pixel of image data into pulse-width data for each pixel, a RAM is provided. The RAM has nine pixel memory regions corresponding to nine pixels a to i for forming, for example, a 3×3 pixel matrix on the output image. In each of the pixel memory regions, there are stored pulse-width data items corresponding to values which are permitted for gray-scale data to have. Pulse-width data items in the pixel memory regions are slightly different from one another among pixel memory regions. Gray-scale data of each pixel of image data and address word composed of gray-scale data of each pixel of image data and coordinates of each pixel in the 3×3 matrix is supplied to the RAM. Thus, pulse-width data corresponding to gray-scale data is read from the pixel memory region so that a laser drive pulse having a pulse width corresponding to the read pulse-width data is generated.Type: GrantFiled: November 18, 1997Date of Patent: April 9, 2002Assignee: Seiko Epson CorporationInventor: Toshiyuki Hattori
-
Patent number: 6353688Abstract: The method identifies a Pixon element, which is a fundamental and indivisible unit of information, and a Pixon basis, which is the set of possible functions from which the Pixon elements are selected. The actual Pixon elements selected from this basis during the reconstruction process represents the smallest number of such units required to fit the data and representing the minimum number of parameters necessary to specify the image. The Pixon kernels can have arbitrary properties (e.g., shape, size, and/or position) as needed to best fit the data.Type: GrantFiled: June 14, 1999Date of Patent: March 5, 2002Assignee: The Regents of the University of CaliforniaInventors: Richard Puetter, Amos Yahil
-
Patent number: 6351571Abstract: A method of analyzing the low- and high-frequency components of a digital image comprising the steps of: providing a rectangular digital image; specifying the size of a rectangular kernel; partitioning the image into nine sub-regions; processing each of said sub-regions using a unique image processing algorithm; combining said processed sub-regions into a resulting image of the low frequency component and of the original image subtracting the low frequency component from the original image to obtain the high frequency component of said original image.Type: GrantFiled: December 14, 1998Date of Patent: February 26, 2002Assignee: Eastman Kodak CompanyInventors: Richard L. VanMetter, Kevin S. Kohm, Xiaohui Wang
-
Publication number: 20020018596Abstract: With respect to two pattern sets obtained on different conditions, a feature-extraction matrix, which maximizes between-class scatter and minimizes within-class scatter, is found respectively. A first feature amount is calculated using one of the feature-extraction matrices. The first feature amount and the two matrices are retained in a referential database. A pattern is determined to a second feature amount—extracted by applying another feature-extraction matrix to a pattern input—by extracting a most similar element out of the first feature amount retained in the referential database.Type: ApplicationFiled: June 5, 2001Publication date: February 14, 2002Inventor: Kenji Nagao
-
Patent number: 6347152Abstract: A multiresolutional filter called a critical point filter is introduced. This filter extracts a maximum, a minimum, and two types of saddle points of pixel intensity for every 2×2 (horizontal×vertical) pixels so that an image of a lower level of resolution is newly generated for every type of a critical point. Using this multiresolutional filter, a source image and a destination image are hierarchized, and source hierarchical images and destination hierarchical images are matched using image characteristics recognized through a filtering operation.Type: GrantFiled: October 23, 2000Date of Patent: February 12, 2002Assignee: Monolith Co., Ltd.Inventors: Yoshihisa Shinagawa, Tosiyasu Laurence Kunii
-
Patent number: 6341182Abstract: A method of analysing an image to obtain an image value. The image comprises a defined array of pixel values. The method comprises (1) determining the highest pixel value in the image; (2) calculating a lower threshold pixel value from the highest pixel value determined in step (1) in accordance with a predetermined algorithm; and (3) obtaining the image value by statistically analysing the pixel values in the image which lie in a range defined by the lower threshold pixel value calculated in step (2).Type: GrantFiled: August 21, 1998Date of Patent: January 22, 2002Assignee: Randox Laboratories Ltd.Inventors: Stephen Peter Fitzgerald, Robert Ivan McConnell, John Victor Lamont
-
Patent number: 6337925Abstract: A method is used for identifying a border in a digital image that is defined by a plurality of pixels, each pixel being defined by a pixel color and a pixel position indicating a location of the pixel in the digital image. The method includes receiving for each pixel a pixel gradient indicating a direction and magnitude of change in color. A user inputs an area of interest that includes at least a portion of the border to be identified. The method then includes estimating information about an edge zone that models the border portion including estimating a position, direction, and width of the edge zone. The position of the edge zone is estimated by calculating a weighted average value of pixel positions of each pixel in the area of interest. The direction of the edge zone is estimated by calculating a weighted average value of pixel gradient direction of each pixel in the area of interest.Type: GrantFiled: May 8, 2000Date of Patent: January 8, 2002Assignee: Adobe Systems IncorporatedInventors: Scott D. Cohen, Gregg Wilensky
-
Patent number: 6332044Abstract: The present invention is a method and apparatus for performing filtering operations on a digital image. One embodiment is enhancement of the contour fidelity of the image represented by a plurality of pixels, by storing pixel values representing a segment of the image in memory; isolating a plurality of regions of the pixels of said image segment to form a plurality of windows including a central pixel and a plurality of pixels surrounding said central pixel. Then comparing at least two of the regions of pixels within the windows with a set of standard pixel patterns for each of the windows and identifying matching pixel patterns within each of the windows.Type: GrantFiled: January 21, 1997Date of Patent: December 18, 2001Assignee: Xerox CorporationInventors: Robert P. Loce, Michael Branciforte
-
Patent number: 6332038Abstract: The present image processing device includes an image obtaining device which obtains an electronic image (still image) or a series of electronic images (moving image), an image extraction processing device which performs image extraction processing for extracting from the image obtained by the image obtaining device a specified object domain an operator wishes to cut out, and a result output device which outputs an image extracted by the image extraction processing device. The image extraction processing device sets two mutually exclusive provisional domains in the obtained image, and, based on separability of the provisional domains, extracts a partial image containing the specified object domain.Type: GrantFiled: April 6, 1999Date of Patent: December 18, 2001Assignee: Sharp Kabushiki KaishaInventors: Ryuji Funayama, Minehiro Konya, Hajime Takezawa
-
Patent number: 6317509Abstract: A method and apparatus detects sheet explosives in computed tomography (CT) data. In particular, sheet-shaped objects such as sheet explosives can be discriminated from other object shapes and detected. The detection includes analyzing a neighborhood of voxels surrounding a test voxel. If the density of the test voxel is sufficiently different from the mean density of the neighboring voxels, then it is concluded that the test voxel is associated with a sheet object. Sheet objects can also be detected by eroding the CT data so as to eliminate voxels associated with thin objects. Remaining objects are then subtracted from the original data, leaving only thin sheet-shaped objects. Erosion of the data can be performed by identifying a neighborhood of voxels surrounding a voxel of interest. If the number of voxels having densities below a predetermined threshold exceeds a predetermined number, then it is assumed that the test voxel is a surface voxel and is removed from the object.Type: GrantFiled: February 11, 1998Date of Patent: November 13, 2001Assignee: Analogic CorporationInventors: Sergey Simanovsky, Ibrahim M. Bechwati, Muzaffer Hiraoglu, Carl R. Crawford
-
Publication number: 20010036296Abstract: One aspect of the invention is a method for locating an unexposed region of film. The method includes the step of illuminating film with a light source while the film has developing chemical applied thereto, the film comprising at least two edges along an x direction perpendicular to a y direction parallel to a surface of the film. The method also includes the step of identifying an unexposed region of the film as a region containing ones of a first plurality of columns of the film, the columns disposed generally in the y direction and captured using at least one sensor operable to capture light reflected from the film, and wherein a representative value for each of the ones of first plurality of columns exceeds a threshold.Type: ApplicationFiled: December 29, 2000Publication date: November 1, 2001Inventors: Robert S. Young, Richard D. Ball, Marc C. Digby, David N. Jones, Philip E. Cannata
-
Publication number: 20010033692Abstract: A process for estimating the noise level of a sequence of images comprises the operations of: producing a local estimate of the noise level of the said images, creating the histogram of the said estimate, deriving at least one parameter of the said histogram, and determining, by calculation or by means of an empirical relation, at least one noise level parameter on the basis of the said at least one parameter derived from the histogram. The corresponding device can be incorporated, for example, in an MPEG-2 encoder, where the parameter identifying the noise level is used for the adjustment of the internal variables of the encoding process.Type: ApplicationFiled: February 16, 2001Publication date: October 25, 2001Inventors: Antonio Maria Borneo, Lanfranco Salinari
-
Patent number: 6304678Abstract: A technique for post-processing decoded compressed images to reduce decoding-related artifacts employs a maximum likelihood estimation of an original image f. The decoded image is modeled as a montage of “flat surfaces” of different intensities, where the number of flat surfaces and their intensities are generally different in different regions of the decoded image. The intensity of each pixel is conditionally adjusted to that of a corresponding flat surface in a window region surrounding the pixel. In a general algorithm, the flat surface model is fitted to the observed image by estimating the model parameters using the “k-means” algorithm and a hierarchical clustering algorithm. A cluster similarity measure (CSM) is used to determine the number of intensity clusters, and hence flat surfaces, in the model of a window region surrounding a pixel of interest. The pixel intensity is adjusted to an estimated value which is the mean intensity of the cluster in which the pixel falls.Type: GrantFiled: May 12, 2000Date of Patent: October 16, 2001Assignee: The Trustees of Boston UniversityInventors: Seungjoon Yang, Yu Hen Hu, Truong Q. Nguyen, Damon L. Tull
-
Patent number: 6295382Abstract: A method and apparatus for adaptive noise filtering within a video graphics circuit includes determining an average intensity for a kernel of a display area. The kernel includes a plurality of pixels arranged in a particular manner, for example a square, a rectangle, etc. Next, a variance for a pixel within the kernel is determined. Based on a relationship between the average intensity and the variance, a signal-to-noise factor is determined. The signal-to-noise factor includes a noise region, a signal region, and an edge region. The pixel within the kernel is then adaptively filtered based on the signal-to-noise factor, the average intensity, and intensity of the pixel.Type: GrantFiled: May 22, 1998Date of Patent: September 25, 2001Assignee: ATI Technologies, Inc.Inventor: Marinko Karanovic
-
Patent number: 6289126Abstract: The boundary of an object is determined based on its optically captured image by: (a) converting a density data of each of a plurality of pixels forming the captured image into a binary signal; (b) converting the binary signal into a neighbor pixel state value so that only the pixels located on the boundary r in the interior of the object have a value corresponding to the binary signals of the surrounding pixels; (c) normalizing the neighbor pixel state value using an absolute chain direction value of a previous chain direction; (d) obtaining a relative chain direction value for a pixel located on the boundary of the object in the image relative to the direction from which the chain has proceeded, with reference to the neighbor pixel state value of the pixel; (e) calculating an absolute chain direction value of a next chain direction on the basis of the relative chain direction value; and (f) repeating the steps (c) to (e) for each pixel located on the boundary to trace the boundary of the object for determinType: GrantFiled: August 7, 1998Date of Patent: September 11, 2001Assignee: Sysmex CorporationInventor: Masaki Ishisaka
-
Patent number: 6285801Abstract: A filter reduces artifacts, such as grid noise and staircase noise, in block-coded digital images with image block boundaries. The type of filtering is determined after an estimation of the image global metrics and local metrics. For areas of the image near grid noise, the filter performs low pass filtering. For image fine details, such as edges and texture, no filtering is performed so that masking is avoided. The filter operates in intra-field mode and uses a fuzzy logic process, pixel deltas, and dual ramp generators to determine the horizontal and vertical length of a processing window surrounding an image block boundary.Type: GrantFiled: May 29, 1998Date of Patent: September 4, 2001Assignee: STMicroelectronics, Inc.Inventors: Massimo Mancuso, Antonio Maria Borneo
-
Patent number: 6282328Abstract: The invention provides machine vision systems and methods for morphological transformation of a source image, e.g., with non-uniform offsets. According to these, pixels from selected “neighborhoods,” or regions, of the source image are rearranged, or unfolded, into respective columns of an intermediate image. For each column in the intermediate image, the method identifies a pixel intensity value of selected rank. Once the pixel intensity value of selected rank is identified for each column, it is retained in a transformation image, e.g., at a location corresponding to center of the respective neighborhood. The foregoing steps are repeated, e.g., along strips or bands, for each neighborhood in the source image.Type: GrantFiled: January 28, 1998Date of Patent: August 28, 2001Assignee: Cognex CorporationInventor: Ranjit P. Desai