Patents Examined by Timothy M. Johnson
  • Patent number: 7840080
    Abstract: An adaptable motion estimator architecture for low bit rate image communication 1) to be compatible with image characteristic and bit rate with a reduced hardware size and 2) to optimize performance of the motion estimator by selectively applying a search method suitable for a low bit rate image characteristic and an encoder performance. The motion estimator multiplexes a previous search window memory data from DRAM and a current macro block data for finding motion vectors to conform to each data processing elements (PE0-PE8) and comparatively detecting MAE (Mean Absolute Error) of each motion vector with a previous frame data and a current frame data to find a motion vector having a least MAE. The motion estimator may be applied to an image phone which requires high encoding efficiency due to small hardware and may be applied to all video encoders conforming to H.261/H.263 and MPEG.
    Type: Grant
    Filed: September 8, 2000
    Date of Patent: November 23, 2010
    Assignee: C & S Technology Co., Ltd.
    Inventor: Young Su Lee
  • Patent number: 6917706
    Abstract: A ruled line extracting apparatus obtains circumscribed rectangles of pixel concatenation regions included in an input pattern, and calculates the most frequent value of their heights. Additionally, the apparatus integrates segments by ignoring a wild card segment, and calculates the most frequent value of height/width of extracted straight lines and segments structuring the straight line. Next, it performs a process for integrating/deleting straight lines using each threshold value based on the highest frequency value. Then, it checks/deletes a straight line according to a distribution of black pixels around the straight line, and recognizes the remaining straight lines as ruled line candidates.
    Type: Grant
    Filed: January 8, 2001
    Date of Patent: July 12, 2005
    Assignee: Fujitsu Limited
    Inventor: Yutaka Katsuyama
  • Patent number: 6912318
    Abstract: A method and system are disclosed for compressing motion image information by means of predictive coding with a high compression ratio while maintaining high image quality. An image in each frame is divided into blocks in advance, and each of all blocks is approximated by (replaced with) a single plane defined by three data representing the magnitude of the intensity of the block, the block-to-block gradient of the intensity in the x direction, and the block-to-block gradient of the intensity in the y direction. Furthermore, corresponding pixels are compared with each other in each block consisting of n×m pixels between frames, and differential information is generated according to the comparison result. Pixels indicated as being not greater than a parameter is treated (deleted) as pixels having no difference between frames. Information compression is performed for pixels indicated as being greater than the parameter.
    Type: Grant
    Filed: December 29, 2000
    Date of Patent: June 28, 2005
    Assignee: Kabushiki Kaisha Office NOA
    Inventors: Noriko Kajiki, Satoshi Tanabe
  • Patent number: 6898326
    Abstract: A transformation unit 102 transforms an entered image to multi-resolution space. A quantizer 105 performs vector quantization on a local pattern of an image of a multi-resolution representation. A perspective-order calculating unit 107 extracts a plurality of code words, positions corresponding thereto and/or angle of rotation and/or scale, and the perspective-order relationship of a plurality of these representative vectors, from the quantized image. An algebraic encoder 108 encodes the input image based upon the extracted information. As a result, there are provided an image processing apparatus and method for asymmetric encoding without motion in three dimensions and extraction of a three-dimensional structure. Further, a transformation unit 1103 transforms an entered image to vector field, and a singularity detector 1104 detects a singularity in the transformed image.
    Type: Grant
    Filed: May 17, 1999
    Date of Patent: May 24, 2005
    Assignee: Canon Kabushiki Kaisha
    Inventor: Teruyoshi Washizawa
  • Patent number: 6873738
    Abstract: A memory apparatus for use with a digital picture signal. The apparatus may comprise a first signal processor for receiving an input digital picture signal and for performing a hierarchical encoding process thereon so as to form hierarchical encoded picture data, a memory for storing the hierarchical encoded picture data from the first signal processor, and a second processor for receiving the hierarchical encoded picture data from the memory and for decoding the received hierarchical encoded picture data in accordance with a hierarchical decoding process to restore the input digital picture signal. The first signal processor, the memory and the second processor being disposed on a common semiconductor substrate.
    Type: Grant
    Filed: September 30, 1996
    Date of Patent: March 29, 2005
    Assignee: Sony Corporation
    Inventor: Tetsujiro Kondo
  • Patent number: 6868178
    Abstract: In a method for processing image data, at least one section of a form (10) is raster scanned and converted into a digital image with image points (22, 24, 26), the color of each image point (22, 24, 26) being represented as tone H, saturation S, and brightness B. Using fuzzy technology, the membership of the image points (22, 24, 26) ant their colors in color categories (UB, SF, BF) is determined by way of membership functions and a gray value GW is allocated to each image point (22, 24, 26) depending on the thus determined color category (UB, SF, BF).
    Type: Grant
    Filed: June 25, 1999
    Date of Patent: March 15, 2005
    Assignee: CGK Computer Gesellschaft Konstanz mbH
    Inventors: Bernhard Frei, Markus Schnitzlein
  • Patent number: 6839457
    Abstract: This invention relates to a bone measurement method for measuring a bone shape, structure and architecture on the basis of the tomographic images of a test bone or joint, in more detail, to a bone measurement method characterized by having a template image extraction step for obtaining a wholly continued template image of a bone inner portion surrounded by a cortical bone from the binary image of a test bone cross section, and separating a cortical bone and a cancellous bone by the product of said template image and said binary image, as a bone measurement method which enables the automatic, high-speed and repeatable separation of a cortical bone and a cancellous bone on the basis of the binary image of a test bone or joint cross section and by which a separated three-dimensional image of the cortical bone portion and the cancellous bone portion is obtained as the bases for a non-invasive analyses of three-dimensional bone structure, bone strength, and the like.
    Type: Grant
    Filed: June 2, 2000
    Date of Patent: January 4, 2005
    Assignee: Teijin Limited
    Inventors: Yoshiaki Azuma, Yoshifumi Harada, Norihiro Yamada, Tsutomu Maeda, Tomohiro Ohta
  • Patent number: 6826294
    Abstract: A block matching motion estimation algorithm uses reduced precision data when comparing pictures in the block to be compressed with pictures in a reference block. This yields a plurality of best match candidates. The average of the best match candidates, excluding those candidates which fall outside a predetermined cluster boundary, is determined and then a second search is performed focused on this average. In this way, the amount of hardware for the block matching algorithm is decreased and the speed of its operation is enhanced.
    Type: Grant
    Filed: March 5, 1999
    Date of Patent: November 30, 2004
    Assignee: Koninklijke Philips Electronics N.V.
    Inventor: Michael Bakhmutsky
  • Patent number: 6826308
    Abstract: A method and apparatus for decoding an image data comprising decoding a coded map signal indicative of a position and shape of a content in a screen inputted for every block of picture signals, decoding coded orthogonal transform coefficients; performing an inverse orthogonal transform of the orthogonal transform coefficient; and deriving a regenerative picture signal on the basis of the results of the inverse orthogonal transform, wherein a two-dimensional orthogonal transform of all the coefficients with respect to blocks located inside of a content among the orthogonal transform coefficients is performed, and the inverse orthogonal transform of only the coefficients contained in the content with respect to the blocks containing a boundary portion of the content is performed.
    Type: Grant
    Filed: April 25, 2003
    Date of Patent: November 30, 2004
    Assignee: Kabushiki Kaisha Toshiba
    Inventors: Noboru Yamaguchi, Toshiaki Watanabe, Kenshi Dachiku, Yoshihiro Kikuchi, Takashi Ida, Takeshi Chujoh
  • Patent number: 6819800
    Abstract: A moving image compression/decompression apparatus and method that uses, for example, a wavelet transform technique in order to improve a compression rate is disclosed. The moving image compression/decompression apparatus includes an A/D converter for converting moving image data into digital data. The digital data is divided into a plurality of level regions using a wavelet transformer. The apparatus also includes a quantizer for quantizing the data that has been transform with a predetermined weight that corresponds to each of the regions. The apparatus also includes an SZT coder for performing a lossless DPCM coding with respect to the data quantized sequentially form a high level region to a low level regionusing a similarity between the level regions based on a predetermined SZT map.
    Type: Grant
    Filed: April 6, 2001
    Date of Patent: November 16, 2004
    Assignee: Rifatron Co., Ltd.
    Inventors: Woo Young Jung, Tae Hoon Kwon
  • Patent number: 6816618
    Abstract: In a system for variable-length coding, when image data of a plurality of documents are compressed and stored in a compression memory, the image data is divided into a plurality of areas. Then, attributes are discriminated for each area, and the result of the attribute discrimination is written to an attribute memory. Attribute counters count the numbers of the discriminated attributes. A variable-length compression section compresses the code data with a compression method in response to the attribute data written to the attribute memory, and the compressed is written to a compression memory.
    Type: Grant
    Filed: March 2, 1999
    Date of Patent: November 9, 2004
    Assignee: Minolta Co., Ltd.
    Inventors: Shoji Imaizumi, Junji Nishigaki, Kenichi Morita
  • Patent number: 6816628
    Abstract: A system and method of adding hyperlinked information to a television broadcast. The broadcast material is analyzed and one or more regions within a frame are identified. Additional information can be associated with a region, and can be transmitted in encoded form, using timing information to identify the frame with which the information is associated. The system comprising a video source and an encoder that produces a transport stream in communication with the video source, an annotation source, a data packet stream generator that produces encoded annotation data packets in communication with the annotation source and the encoder, and a multiplexer system in communication with the encoder and the data packet stream generator. The encoder provides timestamp information to the data packet stream generator and the data packet stream generator synchronizes annotation data from the annotation source with a video signal from the video source in response to the timestamp information.
    Type: Grant
    Filed: October 26, 2000
    Date of Patent: November 9, 2004
    Assignee: GoldPocket Interactive, Inc.
    Inventors: Karen Sarachik, Jon Dakss, V. Michael Bove, Jr., Josh Wachman
  • Patent number: 6810145
    Abstract: A process comprises a step (2) of creating histograms of an image representing the occurrences of the luminance values as a function of these values. A further step compares (5) a histogram of an image T(t) with a histogram of a previous image T(t−1), and is characterized in that the step of comparing (5) consists of calculating, for an occurrence histo13 lum (n,t) corresponding to a luminance value n in the image T(t), the absolute differences between this occurrence and the occurrences histo_lum (n+k, t−1) of the previous image, where K is a natural integer and varies from −K to +K. The process is further characterized by choosing the minimum value Min_Diff (n, t) from among these differences, and is characterized further by in summing (6) the minimum values for the set of luminance values of the histogram, and comparing (7) the sum obtained with a threshold.
    Type: Grant
    Filed: April 4, 2001
    Date of Patent: October 26, 2004
    Assignee: Thomson Licensing, S.A.
    Inventors: Gérard Briand, Jean-Yves Babonneau, Alain Verdier
  • Patent number: 6807308
    Abstract: An image is divided into nonoverlapping tiles, and the tiles are processed in a predefined order. Each tile is processed by applying a predefined family of transform layers to the tile so as to generate successive sets of transform coefficients. The sets of transform coefficients correspond to spatial frequency subbands of the image. The subbands are grouped in accordance with the transform layer that generated them. For one or more respective groups of subbands one or more parameters are generated whose value is indicative of the density of image features in the tile. The tile is classified in accordance with the values of the one or more parameters. Based on the classification, a set of quantization factors for the tile are selected, and then the transform coefficients of the tile are scaled by the selected set of quantization factors to as to generate a set of quantized transform coefficients.
    Type: Grant
    Filed: January 17, 2003
    Date of Patent: October 19, 2004
    Assignee: Zoran Corporation
    Inventors: Charles K. Chui, Hong-Ye Gao, Lefan Zhong
  • Patent number: 6804404
    Abstract: A shape information coder divides the pixels in the shape information into subsets representing different subsamplings of the shape information. Each pixel in each subset is coded with reference to a context which may be derived from reference pixels in more than one of the subsets. The context preferably includes reference pixels located on all sides of the pixel being coded, which leads to efficient coding. The shape information coder may have two or more cascaded stages, each operating as above, with one subset of pixels passed from each stage to be coded in the next stage. The resulting coded shape information is useful in, for example, wavelet coding of pictures.
    Type: Grant
    Filed: March 13, 2003
    Date of Patent: October 12, 2004
    Assignee: Oki Electric Industry Co., Ltd.
    Inventor: Yoshihiro Ueda
  • Patent number: 6801636
    Abstract: An image processing apparatus and method for deriving a digital watermark embedded in an image is provided which has a recognizing unit for recognizing a predetermined characteristic area in the image and a deriving unit for extracting the digital watermark from the image data in the recognized characteristic area.
    Type: Grant
    Filed: January 29, 2001
    Date of Patent: October 5, 2004
    Assignee: Canon Kabushiki Kaisha
    Inventors: Tomochika Murakami, Keiichi Iwamura, Yoshihiro Ishida, Hiromitsu Nishikawa
  • Patent number: 6798927
    Abstract: An image processing apparatus and method capable of decoding and varying at high speed encoded image data having an arbitrary image size by using a memory of a small capacity, and a recording medium storing programs executing such a method are provided. In the image processing apparatus and method, image data encoded in each of block units is decoded in the block unit basis. In order to convert the decoded block scan sequential image data into raster scan sequential image data, the decoded image data is divided in the horizontal direction and converted, and the divided and converted data is varied.
    Type: Grant
    Filed: June 11, 2003
    Date of Patent: September 28, 2004
    Assignee: Canon Kabushiki Kaisha
    Inventors: Masato Kosugi, Hideyuki Rengakuji
  • Patent number: 6798921
    Abstract: Images in a plurality of related images are processed such that a specific region of the image in the first frame is designated and a specific image in the specific region is modified while at the same time the information about the image characteristic quantity of the specific region and what was modified about the specific image are both stored. In the processing of the images in the second and subsequent frames, a similar region that is similar in the image characteristic quantity to the initial specific image is extracted and subjected to the same image modification as has been performed on the specific region of said first frame. Therefore, highly amusing images that have been finished to satisfy the customer's request and other needs can be obtained by simple and efficient operations; as a result, prints of high quality that reproduce such images can be produced in high yield.
    Type: Grant
    Filed: March 16, 1999
    Date of Patent: September 28, 2004
    Assignee: Fuji Photo Film Co., Ltd.
    Inventor: Naoto Kinjo
  • Patent number: 6795565
    Abstract: A method of detecting a block-based watermark is provided. The method detects the location of the watermark within a frame of content data regardless of whether the watermark has been shifted or resized.
    Type: Grant
    Filed: March 8, 2001
    Date of Patent: September 21, 2004
    Assignees: Sony Corporation, Sony Electronics Inc.
    Inventor: Peter David Wendt
  • Patent number: 6795579
    Abstract: A method for recognizing handwritten Chinese characters based on stroke recognition comprises steps of: recognizing handwritten strokes, updating stroke code sequences; retrieving in dictionaries/lexicons at least one corresponding character/phrase entry so as to obtain at least one candidate Chinese character/phrase; dynamically displaying the at least one candidate Chinese character/phrase; jumping to the step of recognizing strokes if it is judged that a next stroke is being written; inputting a displayed Chinese character/phrase into computers as the result of recognition if this character/phase is selected by the user.
    Type: Grant
    Filed: March 14, 2002
    Date of Patent: September 21, 2004
    Assignee: International Business Machines Corporation
    Inventors: Donald T. Tang, Hui Su, Qian Ying Wang