← Back to Patent Explorer
US4850026A Expired / Lapsed US ⏱️ 0.0 years remaining

Chinese multifont recognition system based on accumulable stroke features

Assignee: TELECOMMUNICATION LABORATORIES DIRECTORATE GENERAL TELECOMMUNICATIONS MINISTRY OF COMMUNICATIONS
•
Filing: 1988-07-05
•
Granted: 1989-07-18
•
Expires: 2008-07-05
📄 Original PDF 🌐 Google Patents

Abstract

Abstract A Chinese multifont recognition system, in which a scanner can input Chinese character pattern data into a recognition system to recognize all characters to complete the chinese input operation. The recognition system can recognize various printed fonts and chinese characters in various sizes. The recognition method is arranged that all the Chinese characters to be recognized are to be put in the system; all the features of a Chinese character are extracted, and stored in a feature data base. Later, any input Chinese character will recognized in accordance with the data base. Since it is a technique of extracting the features of a character pattern, the system can recognize the multifont Chinese characters. If it is necessary to expand the number of characters in the data base established originally, a character feature learning system may be used for storing new characters therein.

Inventors

  • 👤 Sor-Shenn Jeng
  • 👤 Gan-How Chang
  • 👤 Kuang-Yao Chang
  • 👤 Jang-Keng Lin
  • 👤 Tsann-Shyong Liu
  • 👤 Rong-Hauh Ju
  • 👤 Hsien-Ju Chang
  • 👤 Yung-Lai Wu
  • 👤 Chih-Yen Yang
  • 👤 Jei-Shyong Huang
  • 👤 Shang-Ju Hsieh
  • 👤 Chun-Chu Wu

Bibliographic Details

Application Number —
Jurisdiction US
Data Source google_patents
Last Synchronized 2026-10-30 21:26
CPC Classifications
G; G06; G06V; G06V30/00; G06V30/10; G06V30/18; G06F; G06F18/00; G06F18/20; G06F18/21

Claims Analysis

Total Claims: 5
Claim 1
Independent Claim 372 words
A Chinese multifont recognition system for recognizing multifont and multi-sized Chinese characters; wherein a printed Chinese character is input into a computer via a scanning means and then data corresponding thereto is processed through a recognition process and by means of a character feature learning system, the features of a character are stored in a recognition system so as to expand the storage capacity of Chinese characters and allow multifont Chinese recognition, said system comprising: a host processor to control the execution state of various modules; a scanner, operatively connected to and controlled by said host processor, which can convert data corresponding to the printed Chinese fonts to be recognized into digital data; a scanner interface card disposed between said scanner and host processor for providing a transmission means between said host processor and said scanner; a monitor connected to said host processor for displaying a resultant character pattern; a printer connected to said host processor for printing said resultant character pattern; a character feature learning system, connected to said host processor which includes a scanner driver, a text object detector, a slicer, a feature extractor, an operation processor, and a feature encoder connected together; said learning system instructing the computer to learn a given number of Chinese characters to be recognized, and to read in the features of every character so as to establish a feature data base, which is to be used later for matching an input Chinese character; a character recognition system, connected to said host processor which includes a pre-processor, a feature extractor, a matching recognizer, a matching processor, and a decoder connected together; said character recognition system enabling the character recognized on a paper to be displayed on said monitor or on said printer; and a memory unit connected to said host processor for storing programs and data; wherein said scanner transmits to said host processor, through a scanner interface card, printed Chinese character pattern data, and then said host processor performs the recognition process through said character recognition system or said character feature learning system; and said memory unit stores the features of the characters and programs to be used as reference data so as to match with the features of a given input Chinese character.
Claim 2
Dependent Claim 372 words
A Chinese multifont recognition system for recognizing multifont and multi-sized Chinese characters; wherein a printed Chinese character is input into a computer via a scanning means and then data corresponding thereto is processed through a recognition process and by means of a character feature learning system, the features of a character are stored in a recognition system so as to expand the storage capacity of Chinese characters and allow multifont Chinese recognition, said system comprising: a host processor to control the execution state of various modules; a scanner, operatively connected to and controlled by said host processor, which can convert data corresponding to the printed Chinese fonts to be recognized into digital data; a scanner interface card disposed between said scanner and host processor for providing a transmission means between said host processor and said scanner; a monitor connected to said host processor for displaying a resultant character pattern; a printer connected to said host processor for printing said resultant character pattern; a character feature learning system, connected to said host processor which includes a scanner driver, a text object detector, a slicer, a feature extractor, an operation processor, and a feature encoder connected together; said learning system instructing the computer to learn a given number of Chinese characters to be recognized, and to read in the features of every character so as to establish a feature data base, which is to be used later for matching an input Chinese character; a character recognition system, connected to said host processor which includes a pre-processor, a feature extractor, a matching recognizer, a matching processor, and a decoder connected together; said character recognition system enabling the character recognized on a paper to be displayed on said monitor or on said printer; and a memory unit connected to said host processor for storing programs and data; wherein said scanner transmits to said host processor, through a scanner interface card, printed Chinese character pattern data, and then said host processor performs the recognition process through said character recognition system or said character feature learning system; and said memory unit stores the features of the characters and programs to be used as reference data so as to match with the features of a given input Chinese character.
Claim 3
Dependent Claim 135 words
A system as claimed in claim 1, wherein said character feature learning system comprises a scanner driver connected to said scanner and said host processor regulates and processes the Chinese data on a paper input to said scanner; and a text object detector connected to said scanner driver and host processor detects the position of every computer character in a line of characters; and a slicer connected to said text object detector and said host processor cuts each input Chinese character into a computer character having the same size; and a feature extractor connected to said slicer and host processor extracts the features of a computer character; and a pattern encoding and features collection means connected to said feature extractor and host processor for converting the character features into an inner code for storage convenience.
Claim 4
Dependent Claim 158 words
A system as claimed in claim 1, wherein said feature extractor includes means which can find out the four-direction strokes of a Chinese character, and then divides a character into m * n lattices so as to recognize the direction of a stroke in each lattice before converting the stroke feature into a symbol, which may be a series of "0" and "1" of digital data to represent the features of the lattices; and each character having m * n lattices; and also learning the features of the other printed character patterns through the aforesaid procedures so as to obtain their features respectively; and during a symbolizing process of the features, accumulative requirement being also considered, and all of the features learned being added up with a logic "OR" so as to obtain the features to recognize Chinese characters with a multifont and in various sizes in order to recognize simultaneously all of the printed Chinese fonts learned.
Claim 5
Dependent Claim 50 words
A system as claimed in claim 1, wherein said matching recognizer is arranged such that the features of a character to be recognized are equally divided into four groups, which are matched and recognized one by one with the features in the character data base using a logic "AND" operation.

References Cited (Backward Citations)

US4177448A Cited
US4365235A Cited
US4516262A Cited
US4524456A Cited
US4573196A Cited
US4545067A Cited
US4607386A Cited
US4685142A Cited
US4684926A Cited
US4758979A Cited

USPTO Annuity & Maintenance Windows

Annuity Stage Window Opens Payment Deadline Surcharge Period Until Estimated Fee Status
3.5 Years (Year 4) 1992-07-17 1993-01-16 1993-07-18 $2000.00 Window Closed / Lapsed
7.5 Years (Year 8) 1996-07-17 1997-01-16 1997-07-18 $3760.00 Window Closed / Lapsed
11.5 Years (Year 12) 2000-07-17 2001-01-16 2001-07-18 $7700.00 Window Closed / Lapsed

Assigned Categories & Tags

No categories assigned to this patent yet.

Assign Category Tag