Browse Prior Art Database

Character Segmentation by Correlation Disappearance

IP.com Disclosure Number: IPCOM000092585D
Original Publication Date: 1966-Dec-01
Included in the Prior Art Database: 2005-Mar-05
Document File: 2 page(s) / 83K

Publishing Venue

IBM

Related People

Baumgartner, RJ: AUTHOR [+2]

Abstract

Segmentation circuitry 10 examines digitized video data entered into recognition shift register 11 to determine when a character is completely scanned. Timing control 12 provides signals to scanner control 13. The latter in turn directs the beam of cathode ray tube 14 in a vertical raster scanning pattern to scan the characters on document 15. Photomultiplier 16 is responsive to light reflected from 15 to generate an analog video signal. This signal is applied to circuit 17 which amplifies, clips and digitizes the analog video data. Such data is entered into 11 under control of 12.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 57% of the total text.

Page 1 of 2

Character Segmentation by Correlation Disappearance

Segmentation circuitry 10 examines digitized video data entered into recognition shift register 11 to determine when a character is completely scanned. Timing control 12 provides signals to scanner control 13. The latter in turn directs the beam of cathode ray tube 14 in a vertical raster scanning pattern to scan the characters on document 15. Photomultiplier 16 is responsive to light reflected from 15 to generate an analog video signal. This signal is applied to circuit 17 which amplifies, clips and digitizes the analog video data. Such data is entered into 11 under control of 12.

Scan counter 18 is advanced by twelve for each vertical scan of the character. Outputs from 18 provide information as to how far scanning progresses into the character space. This information is provided to circuitry 10. Segmentation of touching characters can be accomplished if the presence of left side features of characters can be recognized. Further, a number of characters can be clustered into groups based upon similar left side characteristics. In this example, three left side character features permit clustering. At least more than one character have the features of long left vertical, full curve and half curve. Other characters do not permit clustering and they are identified as having single features. The outputs from these four groups of segmentation circuitry are applied to Or 19. The output from Or 19 is used to signal cha...