GB1338287A - Pattern classifying apparatus - Google Patents

Pattern classifying apparatus

Info

Publication number
GB1338287A
GB1338287A GB5570371A GB5570371A GB1338287A GB 1338287 A GB1338287 A GB 1338287A GB 5570371 A GB5570371 A GB 5570371A GB 5570371 A GB5570371 A GB 5570371A GB 1338287 A GB1338287 A GB 1338287A
Authority
GB
United Kingdom
Prior art keywords
word
trial
signature
class
distance
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired
Application number
GB5570371A
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
International Business Machines Corp
Original Assignee
International Business Machines Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by International Business Machines Corp filed Critical International Business Machines Corp
Publication of GB1338287A publication Critical patent/GB1338287A/en
Expired legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/70Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Theoretical Computer Science (AREA)
  • Medical Informatics (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Character Discrimination (AREA)
  • Collating Specific Patterns (AREA)

Abstract

1338287 Character recognitions systems INTERNATIONAL BUSINESS MACHINES CORP 1 Dec 1971 [7 Dec 1970] 55703/71 Heading G4R In a character or pattern recognition system a signature word representing the pattern is compared with a reference word which may identify the pattern and if the difference or distance between the words exceeds a certain value a class word, from which the reference word is derived, is modified and a further comparison made. The class word is successively modified until a preset difference is obtained, the reference word then indicating the correct class of the character. Signals from the character scanner pass via a pre-processor to a store 210, Fig. 2. From the store groups of signals representing portions of of the character field, i.e. sub-patterns, are passed to respective logic modules 220. There may, for example, be 32 modules. Each module provides a single bit indicative of whether or not a particular feature is present in the corresponding sub-pattern. The bits form a 32-bit signature word representing the character. The signature casues a 6-bit trial class (or position code) word to be generated by a unit 241 which in turn causes unit 243 to generate a trial reference word. The reference word has 32 bits and lies at a distance of at least 16 from each other reference word. The trial reference word and the signature are compared at 244 and cause the trial code word, and hence the trial reference word, to be modified if necessary. When the distance between the compared words is less than 8 the current trial class word is deemed to identify the pattern. This word is passed to store 260 together with the distance value. The class word may be translated into a suitable code, e.g. ASCII, if required. A processor employs the class word and distance value, the latter determining how much reliance may be placed on the accuracy of the identification of the pattern. The logic modules may be binary decision trees, Fig. 4-7 (not shown), or may be structured in a computer program. Units 241-244 may also be structured in a program, Figs. 8, 9 (not shown). In one such structure the bits in each of a number of pairs of fields selected from the signature are compared. Dependent upon the number of mismatches a corresponding 5-bit trial class word is selected and a 32-bit trial reference word constructed from this. The trial reference word and the signature are compared and, dependent on the number of mismatches, the trial class word is modified by addition of a sixth bit, this word now identifying the pattern if the distance between the reference word and the signature is less than 8. Another structure makes use of the fact that most signatures will have a distance 0 or 1 from one of the possible reference words. This structure attempts to form a class word by a procedure involving the comparison of 8-bit bytes of the signature with the initial 8-bits of each reference word. Should the procedure determine that no reference word exists for which the distance is 0 or 1 it is then necessary to compare the signature with all of the reference words.
GB5570371A 1970-12-07 1971-12-01 Pattern classifying apparatus Expired GB1338287A (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US9567670A 1970-12-07 1970-12-07

Publications (1)

Publication Number Publication Date
GB1338287A true GB1338287A (en) 1973-11-21

Family

ID=22253099

Family Applications (1)

Application Number Title Priority Date Filing Date
GB5570371A Expired GB1338287A (en) 1970-12-07 1971-12-01 Pattern classifying apparatus

Country Status (7)

Country Link
US (1) UST966011I4 (en)
AT (1) AT326738B (en)
CH (1) CH534927A (en)
DE (1) DE2159884A1 (en)
FR (1) FR2117165A5 (en)
GB (1) GB1338287A (en)
IT (1) IT941904B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4451929A (en) * 1978-11-10 1984-05-29 Hajime Industries Ltd. Pattern discrimination method
CN111709464A (en) * 2020-06-02 2020-09-25 中科九度(北京)空间信息技术有限责任公司 System and screening method based on data dictionary configuration and seal image recognition technology

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3112093A1 (en) * 1981-03-27 1982-10-07 Eike Prof. Dr.-Ing. 3392 Clausthal-Zellerfeld Mühlenfeld Method for detecting patterns with controlled measurement data selection
US4499596A (en) * 1982-06-28 1985-02-12 International Business Machines Corporation Adaptive facsimile compression using a dynamic extendable decision network
JP2004126841A (en) * 2002-10-01 2004-04-22 Renesas Technology Corp Method for mounting program

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4451929A (en) * 1978-11-10 1984-05-29 Hajime Industries Ltd. Pattern discrimination method
CN111709464A (en) * 2020-06-02 2020-09-25 中科九度(北京)空间信息技术有限责任公司 System and screening method based on data dictionary configuration and seal image recognition technology
CN111709464B (en) * 2020-06-02 2023-06-20 中科九度(北京)空间信息技术有限责任公司 System and screening method based on data dictionary configuration and seal image recognition technology

Also Published As

Publication number Publication date
DE2159884A1 (en) 1972-07-06
IT941904B (en) 1973-03-10
AT326738B (en) 1975-12-29
ATA1038871A (en) 1975-03-15
AU3585971A (en) 1973-05-24
UST966011I4 (en) 1978-01-03
FR2117165A5 (en) 1972-07-21
CH534927A (en) 1973-03-15

Similar Documents

Publication Publication Date Title
KR900018851A (en) Method and apparatus for object direction detection using feedforward neural network
JP2607818B2 (en) Method and apparatus for determining whether a record is stored in a computer system
GB1331987A (en) Preprocessing method and apparatus for pattern recognition
ATE48327T1 (en) METHOD AND APPARATUS FOR ALLOWING THE HOLDER OF A PORTABLE ITEM, IN PARTICULAR A CARD, TO ACCESS A SERVICE FROM AT LEAST ONE AUTHORIZING ENTITY WITH THIS CARD.
DE3273916D1 (en) Methods of data error correction
GB1366009A (en) Reading apparatus
GB1498221A (en) Character recognition unit
CN112988440B (en) System fault prediction method and device, electronic equipment and storage medium
US4003025A (en) Alphabetic character word upper/lower case print convention apparatus and method
CA2006230A1 (en) Method and apparatus for validating character strings
GB1437586A (en) Character recognition system
GB1338287A (en) Pattern classifying apparatus
US20200042422A1 (en) Log analysis method, system, and storage medium
JPH06282686A (en) Method for character recognition
DE69029004D1 (en) Signature verification process
US5177793A (en) Accurate recognition of input patterns
GB1028288A (en) Specimen identification techniques
KR910012982A (en) Specific Data Selection Method for Data Processing Equipment
GB1403816A (en) Apparatus for identifying an unidentified item of data
GB1326141A (en) Raster process for classifying characters
Cuesta et al. Strong laws of large numbers in abstract spaces via Skorohod's Representation Theorem
JP2930996B2 (en) Image recognition method and image recognition device
SU964734A1 (en) Device for reading-out information from associative storage
KR900700973A (en) Character Recognition Device
CN118228063A (en) Message comparison method and computing device

Legal Events

Date Code Title Description
PS Patent sealed [section 19, patents act 1949]
PCNP Patent ceased through non-payment of renewal fee