会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 61. 发明专利
    • PROGRAM GENERATING DEVICE
    • JPH04104324A
    • 1992-04-06
    • JP22097690
    • 1990-08-24
    • FUJI XEROX CO LTD
    • ITONORI KATSUHIKOSHIMIZU NOBORU
    • G06F9/06G05B19/05G06F9/44G06K9/00
    • PURPOSE:To generate a program which a machine can automatically process and which is encoded by separating the picture of the processing procedure of the program into a character area picture and a linear graphic area picture and recognition- processing them to combine recognition results. CONSTITUTION:A picture input device digital-inputs a flow chart where a character and a graphic are mixed and stores it in a picture memory 2. The picture stored in the memory 2 is separated into the character area image composed only by the character and the graphic area picture composed only by the graphic by a character/ graphic separation device 3. When an outline extraction part 43 takes AND between an original picture and a picture obtained by shifting the original picture to left by one picture element and takes AND between the picture of a processed result and the picture obtained by shifting the original picture to left by one picture element and takes AND of pictures obtained by shifting the original picture upward and downward by one picture element, a picture reduced in four directions by one picture element is formed, and an outline is extracted by taking exclusive OR between the picture and the original picture. A chain encoding part 44 raster-scans the picture whose outline is extracted in the extraction part 43 and a black picture element is substituted with a white picture element. When there is the black picture element, chain encoding is executed.
    • 62. 发明专利
    • FEATURE EXTRACTION DEVICE
    • JPH11203407A
    • 1999-07-30
    • JP371498
    • 1998-01-12
    • FUJI XEROX CO LTD
    • RYU TSUGUAKIITONORI KATSUHIKO
    • G06K9/46
    • PROBLEM TO BE SOLVED: To provide a feature extraction device capable of improving the accuracy of sorting characters into a similar characters category. SOLUTION: A character image of one character is read out from an external storage device 1 and stored in memories 2, 3. A horizontal division part 5 in an area division part 4 divides the character image stored in the memory 2 into M sections in a horizontal direction. A longitudinal division part 6 divides the character image stored in the memory 3 into N sections in a longitudinal direction. On the other hand, a scanning range limitation part 7 sets up a scanning range to 1/P of the width of the character and 1/Q of the height of the character. A feature extraction part 8 scans each section divided by the division part 4 within a range limited by the limitation part 7, counts up the number of pixels reached up to an initial character and outputs a feature value based on the count value. Since the scanning range is limited, a feature value independent of the internal structure of a character can be obtained and the sorting of the character into a category of similar characters can be accurately executed.
    • 66. 发明专利
    • BROKEN LINE EXTRACTING DEVICE
    • JPH0320881A
    • 1991-01-29
    • JP15609989
    • 1989-06-19
    • FUJI XEROX CO LTD
    • ITONORI KATSUHIKO
    • G06T7/60
    • PURPOSE:To reduce the number of threshold values to be set without checking the length of a short vector or an adjacent interval, etc., by applying Hough transformation on a straight line including the short vector, and extracting a broken line by observing the mapping distribution of the line. CONSTITUTION:A short vector data memory part 1 selects the vector with length less than a prescribed value out of generated vectors by reading a diagram image, and stores only data with respect to the vector. The prescribed value can be automatically decided corresponding to the diagram image, for example, at the certain percentage of mean of vector length extracted from the targeted diagram image, etc. Thereby, it is possible to set the threshold value judged as the short vector as a value corresponding to the way of write of a graphic. A mapping function arithmetic part 2 performs the computation of a mapping function made as the mapping of a point by extracting the straight line for the short vector stored in the short vector data memory part 1.
    • 67. 发明专利
    • INFORMATION RETRIEVAL DEVICE
    • JP2000099543A
    • 2000-04-07
    • JP27264198
    • 1998-09-28
    • FUJI XEROX CO LTD
    • ITONORI KATSUHIKO
    • G06F17/21G06F17/30
    • PROBLEM TO BE SOLVED: To provide an information retrieval device which integrally deals with document information of different document formats and retrieves a document through the use of relevance shown in the document. SOLUTION: A document format common part 1 previously makes the document format of document information to be common and it is storaged in a document information storage part 2. When a user designates document information and logic structure from a designation device 5 at the time of retrieval, a logic structure extraction part 3 extracts designated logic structure on designated document information. A document information retrieval part 4 retrieves document information stored in the document information storage part 2 with the content of extracted logic structure as a retrieval key. Retrieved document information is stored in a retrieval result storage part 6. Logic structure is extracted by the logic structure extraction part 3 and a document information retrieval part 4 retrieves document information on respective pieces of document information. A retrieval result is stored in a storage device 9 and relevance between document information is displayed on a display device 7 so that it can be grasped.
    • 68. 发明专利
    • PICTURE PROCESSOR
    • JPH0950531A
    • 1997-02-18
    • JP22281495
    • 1995-08-09
    • FUJI XEROX CO LTD
    • ITONORI KATSUHIKO
    • H04N1/387G06T1/00G06T7/60G06T9/20
    • PROBLEM TO BE SOLVED: To provide a picture processor capable of respectively accurately extracting plural pattern information areas included in a picture and executing conversion corresponding to plural color information. SOLUTION: A pattern information boundary detecting means 1 for detecting the boundary of a pattern from a change in pattern information scans an input picture by a mask with specific size, calculates a featured value by the use of the mask and an area separated from the mask by a fixed distance, temporarily stores feature values over the whole picture, compares each stored featured value with a reference value, and when a difference between both the values is more than a certain threshold, decides picture elements corresponding to the mask as a boundary area. A closed area extracting means 2 detects a closed area surrounded by the boundary based upon the boundary information. The inside of the detected closed area is converted into color information by a color conversion means 3. Since a closed area surrounded by a boundary area is extracted and pattern information in the closed area is converted into color information, pattern information even in a boundary area having a complicated shape can also be converted into color information.
    • 69. 发明专利
    • TABLE RECOGNIZER
    • JPH06139404A
    • 1994-05-20
    • JP31086192
    • 1992-10-27
    • FUJI XEROX CO LTD
    • ITONORI KATSUHIKO
    • G06K9/20
    • PURPOSE:To precisely take out the structure of a table even in regard to such tables where the ruled lines are much omitted or the content is partly omitted. CONSTITUTION:A table recognizer recognizes a table image including the characters and ruled lines and contains a character/ruled line separating part 11 which separates the characters from the ruled lines included in a table image, a character block extracting part 12 which extracts the character blocks from the character images separated by the part 11, a character block extending part 13 which extends evenly the edges of each character block based on the mutual position relations of those character blocks extracted by the part 12, and a row extracting part 14 which extracts the rows out of the character blocks based on the position relation among the character blocks extended by the part 13, and a column extracting part 15 which extracts the columns out of the character blocks based on the position relation among the character blocks extended by the part 13 respectively.