Automated text extraction from source drawing and composition into target drawing with translated text placement according to source image analysis

• US 5,813,018 A
• Filed: 03/16/1995
• Issued: 09/22/1998
• Est. Priority Date: 11/27/1991
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd., Hitachi Microcomputer System Limited
• Original Assignee: Hitachi, Ltd., Hitachi Microcomputer System Limited
• Inventors: Nakagawa, Hiroyuki, Kaji, Hiroyuki
• Primary Examiner(s): FETTING, ANTON W
• Application Number:
  US08/405,926
• Time in Patent Office: 1,286 Days
• Field of Search: 395/144-151, 395/769, 395/768, 395/783, 395/793, 395/803, 364/419.14, 382/291, 382/292, 382/176
• US Class Current: 715/236
• CPC Class Codes: G06T 11/60 Editing figures and text; Combining figures or text
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables

Abstract

A document processing method and system for automizing extraction of source language sentences from figures, which is the pre-processing for figure translation, as well as embedding of translated sentences in drawings, which is the postprocessing in figure translation. The pre-processing consists of recognizing the regions included in a source figure, extracting a sentence from each region, and extracting the topological characteristics of the figure, and the post-processing consists of enlarging shrinking each sentence display region according to the change of the sentence length by translation, and generating a figure which preserves the extracted topological characteristics and includes the enlarged/shrinked sentence display regions.

Claims

1. An automated text extraction and drawing composition method, in a translation system for translating a source drawing including texts in a source language into a target drawing including texts in a target language, said method comprising the steps of:

automatically extracting a source language text from at least one box included in said source drawing;

automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

setting constraints on lengths, inclinations, relative positions and connections of line segments of a target drawing to be the same as those of said source drawing;

translating the extracted source language text into a target language text;

determining coordinates of both end points of line segments of said target drawing under the constraints so that the translated target language text can be contained in a target box corresponding to said at least one box; and

embedding said target language text in said target box.

2. An automated text extraction and drawing composition method according to claim 1, further comprising the steps of:

labeling a plurality of boxes in said source drawing;

labeling the extracted source language texts in correspondence with the labeling of said plurality of boxes; and

generating said target drawing based upon said constraints of said source drawing and the labels of said plurality of boxes and said extracted source language texts.

3. An automated text extraction and drawing composition method according to claim 1, further comprising the steps of:

calculating the size of each of said at least one target box for embedding the respective target language text therein;

determining the size and position of line segments of each of said at least one target box, based upon said constraints of said target box; and

embedding each target language text in respective ones of said at least one target box.

4. An automated text extraction and drawing composition method according to claim 1, wherein said extracting step comprises a step of extracting a plurality of source language texts from a plurality of boxes included in said source drawing, with identification of correspondence of each of a plurality of source language texts to each of said plurality of boxes, respectively.

5. An automated text extraction and drawing composition method, in an information processing system for converting a source drawing including texts in a language into a target drawing including texts in said language, said method comprising the steps of:

automatically extracting a source text from at least one box included in said source drawing;

automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

setting constraints on lengths, inclinations, relative positions and connections of line segments of a target drawing to be the same as those of said source drawing;

determining coordinates of both end points of line segments of said target drawing under the constraints so that a target language text can be contained in a target box corresponding to said at least one box, said target text being a corrected text of said source text; and

embedding said target text in said target box.

6. An automated text extraction and drawing composition method according to claim 5, further comprising the steps of:

calculating the size of each of said at least one target box for embedding the respective target text;

determining the size and position of the line segments of each of said at least one target box, based upon said constraints the target box; and

embedding each of said target texts in the respective one of said at least one target box.

7. An automated text extraction and drawing composition method according to claim 5, wherein said extracting step comprises a step of extracting a plurality of source texts from a plurality of boxes included in said source drawing, with identification of correspondence of each of said source texts to each of said plurality of boxes respectively.

8. An automated drawing composition method, in an information processing system for converting a source drawing into a target drawing including texts, said method comprising the steps of:

inputting a target text to be embedded in at least one target box included in said target drawing;

automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

setting constraints on lengths, inclinations, relative positions and connections of line segments of a target drawing to be the same as those of said source drawing;

determining coordinates of both end points of line segments of said target drawing under the constraints so that the target text can be contained in a target box corresponding to said at least one target box; and

embedding said text into said at least one target box.

9. An automated drawing composition method according to claim 8, further comprising the steps of:

calculating the size of said at least one target box for embedding said text therein;

determining the size and position of the line segments of said at least one target box, based upon said constraints; and

embedding said text in said at least one target box.

10. An automated text extraction and drawing composition method, in an information processing system for converting a source drawing including characters into a target drawing including characters, said method comprising the steps of:

reading, as a binary image data, a source drawing including at least one box in which handwritten characters are written;

recognizing said at least one box and said handwritten characters;

automatically extracting the recognized handwritten characters as coded characters to be output as target characters;

automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

setting a constraints on lengths, inclinations, relative positions and connections of line segments of a target drawing to be the same as those of said source drawing;

determining coordinates of both end points of line segments of said target drawing under the constraints so that the target characters can be contained in a target box corresponding to said at least one box; and

embedding said target characters in said at least one target box.

11. An automated text extraction and drawing composition system, in a translation system for translating a source drawing including texts in a source language into a target drawing including texts in a target language, comprising:

means for automatically extracting a source language text from at least one box included in said source drawing;

means for automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

means for representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

means for setting constraints on lengths, inclinations, relative positions and connections of line segments of a target drawing to be the same as those of said source drawing;

means for translating the extracted source language text into a target language text;

means for determining coordinates of both end points of line segments of said target drawing under the constraints so that the translated target language text can be contained in a target box corresponding to said at least one box; and

means for embedding said target language text in said at least one target box.

12. An automated text extraction and drawing composition system according to claim 11, further comprising:

means for labeling a plurality of boxes in said source drawing;

means for labeling the extracted source language texts in correspondence with the labeling of said plurality of boxes; and

means for generating said target drawing based upon said constraints of said source drawing and the labels of said plurality of boxes and said extracted source language texts.

13. An automated text extraction and drawing composition system according to claim 11, further comprising:

means for calculating the size of said at least one target box for embedding the respective target language text;

means for determining the size and position of the line segments of said at least one target box, based upon said constraints the target drawing; and

means for embedding each target language text in the respective ones of said at least one target box.

14. An automated text extraction and drawing composition system according to claim 11, wherein said extracting means comprises means for extracting a plurality of source language texts from a plurality of boxes included in said source drawing, with identification of correspondence of each of said source language texts to each of said boxes respectively.

15. An automated text extraction and drawing composition system, in an information processing system for converting a source drawing including texts in a language into a target drawing including texts in said language, comprising:

means for automatically extracting a source text from at least one box included in said source drawing;

means for automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

means for representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

means for setting constraints on lengths, inclinations, relative positions and connections of a target drawing to be the same as those of said source drawing;

means for determining coordinates of both end points of line segments of said target drawing under the constraints so that a target text can be contained in a target box corresponding to said at least one box, said target text being a corrected text of said source text; and

means for embedding said target text in said target box.

16. An automated text extraction and drawing composition system according to claim 15, wherein said extracting means comprises means for extracting a plurality of source texts from a plurality of boxes included in said source drawing, with identification of correspondence of each of said source texts to each of said boxes respectively.

17. An automated drawing composition system, in an information processing system for converting a source drawing into a target drawing including texts, comprising:

means for inputting a target text to be embedded in at least one target box included in said target drawing;

means for automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

means for representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

means for setting constraints on lengths, inclinations, relative positions and connections of a target drawing to be the same as those of said source drawing;

means for determining coordinates of both end points of line segments of said target drawing under the constraints so that the target text can be contained in a target box corresponding to said at least one box; and

means for embedding said text into said at least one target box.

18. An automated drawing composition system according to claim 17, further comprising:

means for calculating the size of said at least one target box for embedding said text;

means for determining the size and position of the line segments of said at least one target box, based upon said constraints of said target box; and

means for embedding said text in said at least one target box.

19. An automated text extraction and drawing composition system, in an information processing system for converting a source drawing including characters into a target drawing including characters, comprising:

means for reading, as a binary image data, a source drawing including at least one target box in which handwritten characters are written;

means for recognizing said at least one box and said handwritten characters;

means for automatically extracting the recognized handwritten characters as coded characters to be output as target characters;

means for automatically extracting constraints on lengths, inclinations, relative positions and connections of line segments of said source drawing;

means for representing the extracted constraints by formulae indicating relations among coordinates of both end points of said line segments;

means for setting constraints on lengths, inclinations, relative positions and connections of line segments of a target drawing to be the same as those of said source drawing;

means for determining coordinates of both end points of line segments of said target drawing under the constraints so that the target characters can be contained in a target box corresponding to said at least one box; and

means for embedding said target characters in said target box.