Method for locating and circumscribing text areas on a master which may contain text, graphics and/or image areas

• US 4,513,442 A
• Filed: 02/12/1982
• Issued: 04/23/1985
• Est. Priority Date: 02/27/1981
• Status: Expired due to Fees

• Current Assignee: Siemens AG
• Original Assignee: Siemens AG
• Inventors: Scherl, Wolfgang
• Primary Examiner(s): Martin, John C.
• Assistant Examiner(s): NOT, DEFINED
• Application Number:
  US06/348,279
• Time in Patent Office: 1,166 Days
• Field of Search: 382/9, 382/48, 382/49, 382/22
• US Class Current: 382/176
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables

Abstract

A method for locating and circumscribing text areas on a master which may contain text, graphics and/or image areas employs the steps of optoelectronically scanning the master, digitizing the analog signals obtained from the scanning, processing the digital values for forming a plurality of isolated black blocks, and statistically analyzing the coordinates corresponding to the beginning and the end of the black blocks in order to obtain a statistical decision as to whether a particular black block being analyzed represents a line of text.

Claims

1. A method for locating and circumscribing text areas on a master which may contain text, graphics and/or image areas comprising the steps of:

optoelectronically scanning said master for generating analog signals corresponding to an optoelectronic imaging of said master;

converting said analog signals into digital signals by undertaking a decision operation by assigning a first binary number to analog signals representing a white level and assigning a second binary number to analog signals representing a black level;

expanding all points having a black level line-wise by a predetermined multiplier such that a horizontal stroke of predetermined length is respectively generated for each point whereby a black block is generated for each line of text on said master, said black block having a length equal to the length of the line of text plus a predetermined extension length for said black points;

expanding all points remaining after the previous step representing a white level by a predetermined multiplier for expanding said white points in a direction opposite to the expansion of said black points, the expansion of said white points being greater than the expansion of said black points;

extending said black block by a differential length with respect to the original length of the line of text such that a black block extending over the entire line of text is generated;

analyzing said black block with an operator for determining whether horizontal white/black transitions characteristic of text areas exist having a predetermined white column length and a predetermined black column length;

generating a vertical black stroke of predetermined length for every white/black transition such that an isolated black block is generated for each line of text;

listing the left and right extreme coordinates of said black block; and

statistically analyzing said extreme coordinates for said black blocks for determining whether said coordinates are boundaries for a line of text.

2. The method of claim 1 wherein said operator is a window of scanning lines stored in a rolling manner such that a predetermined number of first scanned lines are inverted and a condition for determining the existence of a white/black transition is determined by a column-wise sum formation of the scanned line such that the condition is present when said column sum is equal to 0.

3. The method of claim 1 wherein a counter is provided per output column for said vertical black strokes and comprising the additional steps of reducing said counter by 1 for each line and erasing said counter at the beginning of a new operation and setting said counter to a predetermined length when the operator condition is satisfied indicating a white/black transition, said operator condition being satisfied whenever the value of a respective column counter is greater than 0.

4. The method of claim 1 comprising the additional step of analyzing each output line of vertical black strokes generated by said operator for the existence of black components by the steps of:

assigning a number to each newly beginning black area which number is incremented by 1 for each such newly beginning black area such that the number for each area is continued within an uninterrupted black area within a scanning line;

upon determination of contact between a black area already numbered in a preceding line and a black area to be newly numbered in a line currently being scanned, assigning the numbering of the already numbered black area in the preceding line to the overall black area extending over the two lines;

upon contact of a black area of a currently scanned line with a plurality of black areas of an earlier scanned line, numbering said black area of said newly scanned line with the number of the area of said earlier line disposed farthest to the left;

checking to determine whether numberings of contacting areas disposed further to the right already exist in black areas of the line disposed further to the left and if such condition is met numbering the area in the newly scanned line with the same number;

entering each said number in a table entry with the extreme coordinates for the area associated with said number;

substituting the extreme coordinates for a black area in a subsequent line which contacts the black area having coordinates already entered in said table;

transferring the extreme coordinates of a black area upon termination of said black area to an output list; and

employing said extreme coordinates for said black area as the extreme coordinates for undertaking said statistical analysis.

5. The method of claim 1 wherein said analog signals generated by said optoelectronic scanning are divided into white levels and black levels by analog means.

6. The method of claim 1 wherein said analog signals generated by said optoelectronic imaging are converted into digital values representing white levels and black levels by comparison of said analog signal with a predetermined digital threshold value.

7. The method of claim 6 wherein said threshold value is a constant.

8. The method of claim 7 wherein said threshold value is calculated based on the environment in which the method is employed.

9. The method of claim 8 wherein said threshold value is calculated based on the characteristics of a device employed for said optoelectronic scanning.

10. The method of claim 6 wherein said threshold value is continuously recalculated from said analog signals.