Character recognition apparatus

• US 4,813,078 A
• Filed: 04/29/1988
• Issued: 03/14/1989
• Est. Priority Date: 02/15/1985
• Status: Expired due to Term

• Current Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Original Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Inventors: Shimizu, Masahiro, Takenouchi, Mariko, Fujiwara, Akie
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Couso, Jose L.
• Application Number:
  US07/188,624
• Time in Patent Office: 319 Days
• Field of Search: 382/9, 382/10, 382/16, 382/18-23, 382/25, 382/40, 382/51, 382/56, 358/166, 358/263, 358/282, 358/284
• US Class Current: 382/185
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/182 by coding the contour of the pattern
• CPC Class Codes: G06V 30/184 by analysing segments intersecting the pattern

Abstract

A character recognition apparatus is arranged such that: a character pattern is extracted as a rectangle from the inputted picture image data; the number of picture element points corresponding to the direction code of the boundary point of the character portion is provided as the boundary direction density for each region of the recognition object character pattern for each sub-region provided by the division of character pattern; the scanning operation is performed with respect to the side opposite respectively from four sides of the extracted rectangle; the picture element point which changes from the background to the character portion is defined as a change point; the picture element string number is increased by one each time the change point is detected so as to add the picture string number to each picture element; the number of the picture element point corresponding to the picture element string number is provided as the background density for each region of the recognition object character pattern for each sub-region to be divided; the character recognition is performed using the boundary direction density for each region and the background density for each region.

Claims

1. A character recognition apparatus comprising:

a picture image input portion for inputting picture images including recognition object characters;

a picture image memory portion for storing said inputted picture images;

a character extracting portion for extracting a recognition object character region from said picture image memory portion;

a direction code designating portion for designating a direction code of a picture element point of said recognition object character region through its relationship with picture element points adjacent to said picture element point;

a direction density calculating portion for obtaining, for each direction code, the number of picture element portions having the same direction code using the information of the direction code for each of said picture element points designated by said direction code designating portion;

a change point detection portion for detecting, as at least one change point, a picture element point to be changed from a background portion to a character portion at all times of scanning which is directed from at least one end portion of the recognition object character region to the other end portion thereof and is done in lines on the whole portion of the recognition object character region, and for sequentially ranking the detected change points in order from near said one end portion at every line of the scanning;

a background density calculating portion for obtaining, for each group of ranked change points, the number of picture elements from said one end portion or the n-1st change point to the nth change point wherein n is an integer which is equal to or greater than 1; and

a recognition portion for extracting a candidate character using the number of picture element points obtained by said direction density calculating portion and the number of picture elements obtained by said background density calculating portion.

2. A character recognition apparatus comprising:

a picture image input portion for inputting picture images including recognition object characters;

a picture image memory portion for storing said inputted input images;

a character extracting portion for extracting a recognition object character region from said picture image memory portion;

a direction code designating portion for designating a direction code of a picture element point of said recognition object character region through its relationship with picture element points adjacent to said picture element point;

a direction density calculating portion for dividing said region extracted by said character extracting portion into a first plurality of sub-regions to obtain, for each direction code, the number of picture element portions each to be included in each of said sub-regions with the same direction code using the information of the direction code for each of said picture element points designated by said direction code designating portion;

a change point detection portion for detecting, as at least one change point, a picture element point to be changed from a background portion to a character portion at all times of scanning which is directed from at least one end portion of the recognition object character region to the other end portion thereof and is done in lines on the whole portion of the recognition object character region, and for sequentially ranking the detected change points in order from near said one end portion at every line of the scanning;

a background density calculating portion for dividing said region extracted by said character extracting portion into a second plurality of sub-regions to obtain, for each group of ranked change points, the number of picture elements from said one end portion or the n-1st change point to the nth change point for each of said sub-regions, wherein n is an integer which is equal to or greater than 1; and

a recognition portion for extracting a candidate character using the number of picture element points obtained by said direction density calculating portion for each of said regions and the number of picture elements obtained by said background density calculating portion for each of said regions.

3. A character recognition apparatus in accordance with claim 2, wherein said first sub-regions of said direction density calculating portion for each of said regions and said second sub-regions of said background density calculating portion for each of said regions are the same divided regions.

4. A character recognition apparatus in accordance with claim 2, wherein a recognition object character region to be extracted by said character extracting portion is rectangular in shape.

5. A character recognition apparatus in accordance with claim 2, wherein said recognition object character region extracted by said character extracting portion is divided in its horizontal or vertical direction to respectively determine said first sub-regions of said direction density calculating portion for each of said regions and said second sub-regions of said background density calculating portion for each of said regions.

6. A character recognition apparatus in accordance with claim 2, wherein a picture element point is considered a boundary point to obtain a direction code in said boundary point when the polarity of a picture element point of a character pattern of said recognition object character region is different from the polarity of at least one picture element point selected from among said picture element points which are adjacent to said picture element point.

7. A character recognition apparatus in accordance with claim 2, wherein a change point detection portion scans a recognition object character region from four directions which include:

a direction from left to right, a direction from top to bottom, a direction from right to left, and a direction from bottom to top, so as to detect said change point and sequentially ranks the detected change points in order from closer to the start end portion of said four directions.

8. A character recognition apparatus comprising:

a picture image input portion for inputting picture images including recognition object characters;

a picture image memory portion for storing said inputted picture images;

a character extracting portion for extracting a recognition object character region having a rectangular shape from said picture image memory portion;

a direction code designating portion for designating a picture element point as a boundary point when the polarity of a picture element point of a character pattern of said recognition object character region is different from the polarity of at least one picture element point selected from among eight picture element points adjacent to said picture element point, and for designating a direction code of said boundary point through its relationship with eight picture element points adjacent to said boundary point;

a profile direction density calculating portion for obtaining, for each direction code, the number of boundary points having the same direction code using the information of the direction code for each of said boundary points designated by the direction code designating portion for each sub-region, wherein a region extracted by said character extracting portion is M-divided in a horizontal direction, and is N-divided in a vertical direction;

a change point detection portion for detecting, as at least one change point, a picture element point to be changed from background portion to a character portion at all times of scanning which is directed from at least one end portion of the recognition object character region to the other end portion thereof in four directions including a direction from right to left, a direction from bottom to top and vice versa and is done in lines on the whole portion of the recognition object character region, and for sequentially ranking the detected change points in order from near said one end portion at every line of the scanning;

a background density calculating portion for obtaining, for each group of ranked change points, the number of picture elements from the start end portion or the n-1st change point to the nth change point for each of said sub-regions, wherein n is an integer which is equal to or greater than 1, wherein a region extracted by said character extracting portion is M-divided in said horizontal direction, and is N-divided in said vertical direction; and

a recognition portion for extracting a candidate character using the number of picture element points obtained by said profile direction density calculating portion for each of said regions and the number of picture elements obtained by said background density calculating portion for each of said regions.