KR101498546B1 - System and method for restoring digital documents - Google Patents

Abstract

The present invention relates to a document digital restoration system and method, and a document digital restoration method according to the present invention comprises the steps of classifying a print block, a handwritten block and a picture block from a document image by analyzing a document image; Generating digital ink data for a handwritten block that fails in character recognition when the character recognition for the handwritten block fails, and generating digital ink data for the handwritten block, Performing character recognition on each of the print block and the handwritten block, including performing character recognition using ink data; And a step of generating a document for the document image based on the character recognition result and the picture block in the step of performing the character recognition. In addition, not only the character recognition rate for the document image can be increased, It is possible to minimize the inclusion and enhance the ease of editing.

Description

[0001] SYSTEM AND METHOD FOR RESTORING DIGITAL DOCUMENTS [0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document digital restoration system and method, and more particularly, to a document digital restoration system and method for restoring a document image obtained by mixing a printed matter, a writing body, and a picture into a document.

Recently, electronic devices having cameras such as smart phones, tablet PCs, and notebooks have been increasing in popularity. In addition to technologies for retrieving characters or pictures from scanned images and restoring them as digital documents, offline documents such as minutes, A demand for a technique of photographing a document including a printed matter such as a signboard on the street or a menu plate of a restaurant and / or a handwritten character and storing the image as an image or storing the captured image as a character file, Is increasing.

The OCR (Optical Character Reader) technology that recognizes a document or image and converts it into text and restores it as a digital document has a high recognition rate for printed characters. However, in the case of off-line handwritten characters, the character type is not standardized, so the difficulty of character recognition is very high, and the character recognition rate through OCR technology is also very low. Therefore, in the case of a document image with handwritten characters, the handwritten character is treated as an image and edited. However, many of such character recognition methods are inconvenient.
[Related Technical Literature]
A handwriting recognition system and a handwriting recognition method using the same (Korean Patent Application No. 10-2011-0089500)

SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to provide a document digital restoration system and method capable of enhancing a character recognition rate for a document image in which a printed character, a cursive character, and a picture are mixed.

Another object of the present invention is to provide a document digital restoration system and method which can enhance the easiness of editing by minimizing the inclusion of an image in a character recognition result of a document image.

The problems of the present invention are not limited to the above-mentioned problems, and other problems not mentioned can be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention, there is provided a document digital restoration system for analyzing a document image and analyzing the document image to classify a print block, a handwritten block, and a picture block from a document image, A handwritten character recognition unit for performing character recognition on a cursive block, and a handwritten character recognition unit for generating digital ink data for a handwritten block that fails in character recognition if character recognition fails A character recognition section for performing character recognition for each of the print block and the handwritten block, the character recognition section including a digital ink data generation section and a digital ink data character recognition section for performing character recognition using digital ink data; Based on the character recognition result and the picture block, Generating a document to generate a document to be characterized in that it comprises a.

According to another aspect of the present invention, there is further provided an input unit for inputting a document image, wherein the input unit is a scanner or a camera.

According to another aspect of the present invention, a print character recognition unit transmits a print block which has failed recognition of a print block to a handwritten character recognition unit or a digital ink data generation unit, and the handwritten character recognition unit recognizes, Character recognition is further performed.

According to still another aspect of the present invention, a handwritten character recognition unit transmits a handwritable block or a print block, which has failed recognition of a print block received from a handwritten block or a printed character recognition unit, to a digital ink data generation unit, And further generates digital ink data for the print block received from the recognition section, the handwritten block received from the handwritten character recognition section, or the print block.

According to still another aspect of the present invention, there is provided a printing apparatus, including a printed character recognizing unit, a handwritten character recognizing unit, a digital ink data generating unit, and a digital ink data character recognizing unit, And the ink data character recognizing unit are sequentially driven in this order.

According to another aspect of the present invention, a printed character recognition unit, a handwritten character recognition unit, a digital ink data generation unit, and a digital ink data character recognition unit are driven in parallel.

According to another aspect of the present invention, the printed character recognizing unit and the handwritten character recognizing unit are selectively driven.

According to another aspect of the present invention, the print character recognition unit, the handwritten character recognition unit, and the digital ink data character recognition unit are characterized by using a preprocessing process, a feature extraction process, a recognition process, and a post-process process for character recognition .

According to another aspect of the present invention, the digital ink data character recognition unit transmits digital ink data to the document generation unit when the character recognition fails.

According to an aspect of the present invention, there is provided a digital document restoration method comprising: analyzing a document image to classify a print block, a handwritten block, and a picture block from a document image; Generating digital ink data for a handwritten block that fails in character recognition when the character recognition for the handwritten block fails, and generating digital ink data for the handwritten block, Performing character recognition on each of the print block and the handwritten block, including performing character recognition using ink data; And generating a document for the document image based on the character recognition result and the picture block in the step of performing character recognition.

According to another aspect of the present invention, the step of performing character recognition on the handwritten block further includes the step of performing character recognition on the print block that fails to recognize the character when the character recognition on the print block fails do.

According to another aspect of the present invention, the step of generating digital ink data may include a step of performing character recognition on a print block that fails to recognize a character when the character recognition on the print block fails, And performing character recognition on a handwritten block that fails in character recognition if the character recognition fails.

According to another aspect of the present invention, there is provided a method for performing character recognition on a print block, performing character recognition on a cursive block, generating digital ink data, and performing character recognition using digital ink data Is performed in parallel.

According to an aspect of the present invention, there is provided a computer-readable medium for analyzing a document image to classify a print block, a handwritten block, and a picture block from a document image, The character recognition is performed on the cursive block, the digital ink data on the failed cursive block is generated when the character recognition on the cursive block fails, and the character recognition is performed using the digital ink data To perform character recognition for each of the print block and the handwritten block and to generate a document for the document image based on the character recognition result and the picture block in the step of performing character recognition .

The details of other embodiments are included in the detailed description and drawings.

The present invention can provide a document digital restoration system and method capable of enhancing a character recognition rate for a document image in which a printed character, a cursive character, and a picture are mixed.

The present invention can provide a document digital restoration system and method that can increase the easiness of editing by minimizing the inclusion of an image in a character recognition result on a document image.

The effects according to the present invention are not limited by the contents exemplified above, and more various effects are included in the specification.

1 is a flowchart of a digital document restoration method according to an embodiment of the present invention.
2 is a schematic diagram of a document digital restoration system in accordance with an embodiment of the present invention.
3 is an example of a document image used in the document digital restoration system and method of the present invention.
4 is a flowchart illustrating a character recognition step of the digital document restoration method according to an embodiment of the present invention.
5 is a schematic view of a character recognition unit of a document digital restoration system according to an embodiment of the present invention.
6 is a schematic view of a printed character recognition unit of a character recognition unit of a document digital restoration system according to an embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention and the manner of achieving them will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims.

Like reference numerals refer to like elements throughout the specification.

Each block of the accompanying block diagrams and combinations of the steps of the flowcharts may be performed by algorithms or computer program instructions comprised of firmware, software, or hardware. These algorithms or computer program instructions may be embedded in a processor of a general purpose computer, special purpose computer, or other programmable digital signal processing device, so that the instructions that are executed by a processor of a computer or other programmable data processing apparatus Generate means for performing the functions described in each block or flowchart of the block diagram. These algorithms or computer program instructions may also be stored in a computer usable or computer readable memory capable of directing a computer or other programmable data processing apparatus to implement a function in a particular manner, It is also possible for instructions stored in a possible memory to produce a manufacturing item containing instruction means for performing the function described in each block or flowchart of each block diagram. Computer program instructions may also be stored on a computer or other programmable data processing equipment so that a series of operating steps may be performed on a computer or other programmable data processing equipment to create a computer- It is also possible that the instructions that perform the processing equipment provide the steps for executing the functions described in each block of the block diagram and at each step of the flowchart.

Also, each block or each step may represent a module, segment, or portion of code that includes one or more executable instructions for executing the specified logical function (s). It should also be noted that in some alternative embodiments, the functions mentioned in the blocks or steps may occur out of order. For example, two blocks or steps shown in succession may in fact be performed substantially concurrently, or the blocks or steps may sometimes be performed in reverse order according to the corresponding function.

It is to be understood that each of the features of the various embodiments of the present invention may be combined or combined with each other partially or entirely and technically various interlocking and driving is possible as will be appreciated by those skilled in the art, It may be possible to cooperate with each other in association.

Various embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

1 is a flowchart of a digital document restoration method according to an embodiment of the present invention. 2 is a schematic diagram of a document digital restoration system in accordance with an embodiment of the present invention. Referring to FIG. 2, the document digital restoration system 100 includes a document structure analysis unit 110, a character recognition unit 120, and a document generation unit 130. 1 and 2 for a document digital restoration method and a document digital restoration system 100 will be described below in order to facilitate understanding of the document digital restoration method and the document digital restoration system 100 according to an embodiment of the present invention. .

First, referring to FIG. 1, a document image is analyzed to classify a print block, a cursive block, and a picture block from a document image (S10). Step S10 is performed by the document structure analysis unit 110 of the document digital restoration system 100. [

The document structure analyzing unit 110 analyzes the document image and classifies characters and pictures included in the document image into a print block, a cursive block, and a picture block. The document image may be an image scanned through a scanner, an image captured through a camera, or an image read from an image file, and the document image format may be a variety of image formats such as BMP, JPEG, and PNG. In addition, the document image may have various color depths, such as color images, monochrome images, and the like. Although not shown in FIG. 2, the document digital restoration system 100 may further include a memory, the document image may be stored in a memory, and the document structure analysis unit 110 may access the document image stored in the memory Process the document image.

Although not shown in FIG. 2, the document digital restoration system 100 may include an input unit for receiving a document image. As an input unit, a device for image photographing such as a scanner or a camera may be used.

Hereinafter, the document image and document structure analyzing unit 110 will be referred to together with FIG. 3 for a more detailed description.

3 is an example of a document image used in the document digital restoration system and method of the present invention.

The document image 200 includes a print block 210, a cursive block 220, and a picture block 230. As shown in FIG. 3, the print block 210 refers to a block of characters included in the document image 200 and edited by a word processor or the like. As shown in FIG. 3, the cursive block 220 refers to a block of characters included in the document image 200 in the form manually generated by a person. The picture block 230 refers to a block of pictures or pictures contained in the document image 200, as shown in FIG.

The document structure analysis unit 110 identifies and classifies the print block 210, the handwritten block 220, and the picture block 230 included in the document image 200. The document structure analysis unit 110 analyzes the document image 200 and divides the document image 200 into a print block 210, a cursive block 220, and a picture block 230. In the case of the print block 210 and the handwritten block 220, since the character recognition for the characters included in the corresponding block is performed for restoring the document, the document structure analyzing unit 110 analyzes the print block 210 and the cursive block 220 to the character recognition unit 120. In the case of the picture block 230, the document structure analyzing unit 110 transmits the picture block 230 to the document generating unit 130 because it is possible to restore the document without performing a separate character recognition.

1 and 2, character recognition is performed on each of the print block 210 and the cursive block 220 using the digital ink data, based on the classification of the document structure analyzing unit 110 (S20). Step S20 is performed by the character recognition unit 120 of the document digital restoration system 100. [

The character recognition unit 120 performs character recognition on the print block 210 and the cursive block 220, respectively. 4 and 5 together for a more detailed description of the character recognition unit 120. [

4 is a flowchart illustrating a character recognition step of the digital document restoration method according to an embodiment of the present invention. 5 is a schematic view of a character recognition unit of a document digital restoration system according to an embodiment of the present invention. 5, the character recognition unit 120 includes a print character recognition unit 121, a handwritten character recognition unit 122, a digital ink data generation unit 123, and a digital ink data character recognition unit 124). Hereinafter, in order to understand the character recognizing step of the document digital restoring method and the character recognizing part 120 of the document digital restoring system 100 according to an embodiment of the present invention, And the character recognition unit 120 of the document digital restoration system 100 will be described with reference to FIGS. 4 and 5. FIG.

Referring to FIG. 4, character recognition is performed on the print block 210 (S21). Step S21 is performed by the printed character recognition unit 121 of the character recognition unit 120 of the document digital restoration system 100.

The print character recognition unit 121 performs character recognition on the print block 210 from the document structure analysis unit 110. The print character recognizing unit 121 receives the print block 210 from the document structure analyzing unit 110 and performs character recognition on the print character included in the print block 210. [ For a more detailed description of the character recognizing process of the printed character recognition unit 121, refer to Fig.

6 is a schematic view of a printed character recognition unit of a character recognition unit of a document digital restoration system according to an embodiment of the present invention. The printed character recognition unit 121 includes a preprocessing unit 126, a feature extraction unit 127, a recognition unit 128, and a post-processing unit 129.

The preprocessing of the characters included in the print block 210 is performed by the preprocessing unit 126 of the printed character recognition unit 121 so that the printed character recognition unit 121 recognizes the characters of the printed character . The preprocessing unit 126 corrects inclination of characters included in the print block 210, removes noise such as dirt contained in the print block 210, and determines whether the characters included in the print block 210 are color images A character segmentation process is performed on the characters included in the print block 210. The character segmentation process includes the steps of performing line segmentation, word segmentation, character segmentation, To perform various preprocessing operations.

The feature extraction unit 127 of the printed character recognition unit 121 extracts the characteristics of characters included in the preprinted print block 210. [

The recognition unit 128 of the printed character recognition unit 121 uses various methods such as a structural analysis, a neural network method, and a statistical method based on the characteristics of the characters extracted by the feature extraction unit 127 And generates a character code such as ASCII for the characters included in the print block 210. [

The post-processing unit 129 of the printed character recognizing unit 121 calibrates the erroneously recognized character among the characters recognized by the recognizing unit 128 by correct character using language modeling such as dictionary (lexicon).

Although not shown in Fig. 6, the printed character recognition unit 121 may further include a recognition result evaluation unit. The recognition result evaluating unit calculates the degree of assurance of the character recognition result in the printed character recognition unit (121). For example, the recognition result evaluating unit can calculate the degree of confidence in the character recognition result by numerical calculation, and can determine that recognition is successful if the calculation result is greater than or equal to the predetermined threshold value, and can be determined to be the recognition failure if the calculation result is less than the threshold value.

The printed character recognition unit 121 transmits the character recognition result to the handwritten character recognition unit 122, the digital ink data generation unit 123 or the document generation unit 130 based on the evaluation result of the recognition result evaluation unit. If the evaluation result by the recognition result evaluation unit is successful, the recognition result by the printed matter character recognition unit 121 is transmitted to the document generation unit 130. [ If the evaluation result by the recognition result evaluating unit is a recognition failure, the print block 210 which has failed to be recognized is sent to the handwritten character recognizing unit 122 or the digital ink data generating unit 123. The failure of the character recognizing unit 121 to recognize characters may mean that the characters of the characters included in the print block 210 are different from those of the standard printed characters. Therefore, in order to perform additional character recognition on the print block 210 that failed to recognize the character by the print character recognizing unit 121, the print character recognizing unit 121 recognizes the print block 210 that failed to recognize as a handwritten character recognition To the digital ink data generation unit 123 or the digital ink data generation unit 123. [ In addition, as described above, since the characters included in the print block 210 can be line segmented, word segmented, and character segmented, the printed character recognition unit 121 recognizes the segmented line, segmented word, And the character included in the print block 210 in units of divided lines, divided words, or divided characters is determined by the handwritten character recognition unit 122 or the digital ink data generation unit 123 ).

The printed character recognition unit 121 can be selectively driven. That is, the print character recognition unit 121 may be driven to always perform character recognition on the print block 210, or may be driven to selectively perform character recognition on the print block 210. [ When the print character recognition unit 121 is set to be always driven, character recognition for the print block 210 is performed by the print character recognition unit 121 as described above. However, when the print character recognition unit 121 is selectively driven The print block 210 may be transmitted from the document structure analyzing unit 110 to the handwritten character recognizing unit 122 or the digital ink data generating unit 123 when the print character recognizing unit 121 is set not to be driven .

4 and 5, character recognition is performed on the handwritten block 220 (S22). Step S22 is performed by the handwritten character recognition unit 122 of the character recognition unit 120 of the document digital restoration system 100. [

The handwritten character recognition unit 122 performs character recognition on the handwritten block 220 from the document structure analysis unit 110. [ The handwritten character recognition unit 122 receives the handwritten block 220 from the document structure analysis unit 110 and performs character recognition on handwritten characters included in the handwritten block 220. [ When the character recognition of the print block 210 fails by the print character recognizing unit 121 as described above, the handwritten character recognizing unit 122 receives the print block 210 which has failed to recognize the character, Character recognition may be performed by considering the printed character included in the block 210 as a cursive character. When the printed character recognition unit 121 is set not to be driven as described above, the handwritten character recognition unit 122 receives the printed material block 210 from the document structure analysis unit 110, Character recognition may also be performed on the printed characters included in the text.

The handwritten character recognition unit 122 may include a preprocessor, a feature extraction unit, a recognition unit, and a post-processing unit. The handwritten character recognizing unit 122 includes a preprocessing unit 130 for performing various preprocessing such as skew correction for characters included in the handwritten block 220, noise elimination, binarization to a monochrome image, line segmentation, word segmentation, A feature extracting unit for extracting a feature of characters included in the handwritten block 220 that has been preprocessed, a character extracting unit for extracting characters corresponding to characters included in the handwritten block 220 using various methods based on the extracted characters, A recognition unit for generating a code, and a post-processing unit for correcting the erroneous character using language modeling. The preprocessing unit, the feature extraction unit, the recognition unit, and the post-processing unit of the handwritten character recognition unit 122 are the same as the preprocessing unit 121, except that the processing object is the characters included in the handwritten block 220, The feature extracting unit 127, the recognizing unit 128, and the post-processing unit 129 according to an embodiment of the present invention.

The handwritten character recognition unit 122 may further include a recognition result evaluation unit. The recognition result evaluation unit calculates the degree of certainty of the character recognition result in the handwritten character recognition unit (122). For example, the recognition result evaluating unit can calculate the degree of confidence in the character recognition result by numerical calculation, and can determine that recognition is successful if the calculation result is greater than or equal to the predetermined threshold value, and can be determined to be the recognition failure if the calculation result is less than the threshold value.

The handwritten character recognition unit 122 transmits the character recognition result to the digital ink data generation unit 123 or the document generation unit 130 based on the evaluation result of the recognition result evaluation unit. If the evaluation result by the recognition result evaluating unit is successful, the recognition result by the handwritten character recognizing unit 122 is transmitted to the document generating unit 130. If the evaluation result by the recognition result evaluating unit is a recognition failure, the handwritten block 220 which has failed to be recognized is transmitted to the digital ink data generation unit 123. [ The failure of the recognition of the character by the handwritten character recognition unit 122 may mean that the characters of the characters included in the cursive block 220 are different from the general shape. Therefore, in order to perform additional character recognition on the handwritten block 220 that has failed to recognize the character by the handwritten character recognition unit 122, the handwritten character recognition unit 122 recognizes the handwritten block 220, To the generating unit 123. As described above, since the characters included in the cursive block 220 can be line segmented, word segmented, and character segmented, the handwritten character recognition unit 122 recognizes the segmented line, segmented word, And can transmit characters included in the cursive block 220 to the digital ink data generation unit 123 in units of divided lines, divided words, or divided characters.

The handwritten character recognition unit 122 can be selectively driven. That is, the handwritten character recognizing unit 122 recognizes the characters in the handwritten block 220, and recognizes that the printed character recognizing unit 121 is not operated and the printed character recognizing unit 121 is not recognized, The character recognition for the handwritten block 220 and the failure to recognize the printed character recognition unit 121 and the printed character recognition unit 121 As shown in FIG. Therefore, when the handwritten character recognition unit 122 is set to be always driven, character recognition for the handwritten block 220 and / or the print block 210 is performed by the handwritten character recognition unit 122 as described above. However, The handwritable block 220 and / or the print block 210 are transmitted to the digital ink data generation unit 123 when the character recognition unit 122 is selectively driven so that the handwritten character recognition unit 122 is not driven .

Then, when the character recognition of the handwritten block 220 fails, the digital ink data for the handwritten block 220 that failed to recognize the character is generated (S23). Step S23 is performed by the digital ink data generation unit 123 of the character recognition unit 120 of the document digital restoration system 100. [

The digital ink data generation unit 123 generates digital ink data for the handwritten block 220 that failed to recognize the character when the handwritten character recognition unit 122 fails to recognize the character for the handwritten block 220. [ That is, the digital ink data generation unit 123 restores the locus of the characters included in the cursive block 220 to generate digital ink data. Digital ink data refers to data generated by a user in a process of handwriting using an input means such as a hand or a stylus on an input device such as a digitizer or a touch screen display, Quot; refers to data generated on the basis of a stroke according to the stroke. A stroke consists of a set of points defined by spatial coordinates. The stroke may be, but is not limited to, a set of points arranged in time sequence along the path of the input from when the user applies the input until the input is released. The digital ink data generation unit 123 restores the dynamic stroke from the static image such as the characters included in the cursive block 220 and generates digital ink data.

The digital ink data generation unit 123 may further include an evaluation unit. The evaluation unit calculates the degree of confidence in the digital ink data generation result in the digital ink data generation unit (123). For example, the evaluating unit can calculate the degree of confidence in the digital ink data generation result by numerically calculating it. If the calculation result is greater than or equal to the predetermined threshold value, it is determined that the generation is successful. If the calculation result is less than the threshold value,

The digital ink data generation unit 123 transmits the digital ink data or image to the digital ink data character recognition unit 124 or the document generation unit 130 based on the evaluation result of the evaluation unit. If the evaluation result by the evaluation unit is successful, the digital ink data generation unit 123 transfers the digital ink data to the digital ink data character recognition unit 124. The character included in the handwritten block 220 which has failed to be generated is processed as an image and the character included in the handwritten block 220 or the handwritten block 220 is converted into an image And transmits it to the document generation unit 130.

As described above, the handwritten character recognition unit 122 is selectively driven so that the handwritten character recognition unit 122 can be set not to be driven. In this case, the digital ink data generation unit 123 may receive the handwritten block 220 from the document structure analysis unit 110 and generate digital ink data for the handwritten block 220.

When the handwritten character recognizing unit 122 is set not to be driven, the printed character block 210 failed to recognize the character in the printed character recognizing unit 121 is transferred to the handwritten character recognizing unit 122 as described above But may be transmitted to the digital ink data generation unit 123. In this case, the digital ink data generation unit 123 may receive the print block 210 that failed in character recognition in the print character recognition unit 121, and may generate digital ink data for the print block 210.

Also, the printed character recognition unit 121 may be selectively driven so that the printed character recognition unit 121 is not driven. Accordingly, when both the print character recognition unit 121 and the cursive character recognition unit 122 are set not to be driven, both the print block 210 and the handwritten block 220 included in the document image 200 are displayed in the document structure And may be transmitted from the analysis unit 110 to the digital ink data generation unit 123. In this case, the digital ink data generation unit 123 receives the print block 210 and the handwritten block 220 from the document structure analysis unit 110 and generates digital data for the print block 210 and the handwritten block 220 Ink data can be generated.

On the other hand, when the character recognition unit 121 fails to recognize the characters included in the print block 210, the print character recognition unit 121 outputs the print block 210, which has failed to recognize the character, (123). In this case, the digital ink data generation unit 123 may receive the print block 210 that failed in character recognition in the print character recognition unit 121, and may generate digital ink data for the print block 210.

Then, character recognition is performed using the digital ink data (S24). Step S24 is performed by the digital ink data character recognition unit 124 of the character recognition unit 120 of the document digital restoration system 100. [

The digital ink data character recognition unit 124 performs character recognition using the digital ink data from the digital ink data generation unit 123. [ The digital ink data character recognition unit 124 receives the digital ink data from the digital ink data generation unit 123 and performs character recognition based on the strokes included in the digital ink data.

The digital ink data character recognizing unit 124 may include a preprocessing unit, a feature extracting unit, a recognizing unit, and a post-processing unit. The digital ink data character recognizing unit 124 includes a preprocessing unit for performing various preprocessing such as calibration, size normalization and the like for the strokes included in the digital ink data, a feature extracting unit for extracting the features of the preprocessed strokes, A recognition unit for generating a character code for the characters using various methods based on the recognition result, and a post-processing unit for correcting the erroneous character using language modeling. The preprocessing unit, the feature extracting unit, the recognizing unit, and the post-processing unit of the digital ink data character recognizing unit 124 are the same as the preprocessing unit 126, The feature extracting unit 127, the recognizing unit 128, and the post-processing unit 129 according to an embodiment of the present invention.

The digital ink data character recognition unit 124 may further include a recognition result evaluation unit. The recognition result evaluation unit calculates the degree of assurance of the character recognition result in the digital ink data character recognition unit (124). For example, the recognition result evaluating unit can calculate the degree of confidence in the character recognition result by numerical calculation, and can determine that recognition is successful if the calculation result is greater than or equal to the predetermined threshold value, and can be determined to be the recognition failure if the calculation result is less than the threshold value.

The digital ink data character recognition unit 124 transmits the character recognition result to the document generation unit 130 based on the evaluation result of the recognition result evaluation unit. If the evaluation result by the recognition result evaluation unit is successful, the recognition result by the digital ink data character recognition unit 124 is transmitted to the document generation unit 130. [ If the evaluation result by the recognition result evaluation unit is a recognition failure, the digital ink data is transmitted to the document generation unit 130 for recognition.

Performing character recognition for the print block 210, performing character recognition for the cursive block 220, generating digital ink data, and performing character recognition using digital ink data, . That is, character recognition for the print block 210 by the printed character recognition unit 121 is performed, character recognition for the handwritten block 220 by the handwritten character recognition unit 122 is performed, The digital ink data generating unit 123 generates digital ink data for the pen body block 220 that failed in character recognition and then the digital ink data character recognizing unit 124 recognizes the character using the digital ink data .

On the other hand, performing character recognition for the print block 210, performing character recognition for the cursive block 220, generating digital ink data, and performing character recognition using digital ink data Can be performed in parallel. That is, character recognition for the print block 210 is performed by the print character recognition unit 121, character recognition for the handwritten block 220 by the handwritten character recognition unit 122 is performed, The digital ink data for the handwritten block 220 that failed in character recognition by the digital ink data character recognition unit 124 is generated and the character recognition using the digital ink data by the digital ink data character recognition unit 124 is performed simultaneously .

Referring again to FIGS. 1 and 2, a document for the document image 200 is generated based on the character recognition result and the image block 230 (S30). Step S30 is performed by the document generation unit 130 of the document digital restoration system 100. [

The document generation unit 130 generates character recognition results from the print character recognition unit 121, the handwritten character recognition unit 122 and the digital ink data character recognition unit 124 of the character recognition unit 120, And generates a document for the document image 200 based on the image block 230 from the image processor 110. The document generation unit 130 generates a character recognition result for the printed material block 210 from the printed character recognition unit 121 of the character recognition unit 120 and a character recognition result from the written character recognition unit 122 of the character recognition unit 120 A character recognition result for the cursive block 220 and a character recognition result for the digital ink data from the digital ink data character recognition unit 124 of the character recognition unit 120. The character recognition result can be used as text have. The document generation unit 130 receives the image block 230 of the document image 200 classified by the document structure analysis unit 110. If the character recognition unit 120 fails in character recognition, The block 130 receives the digital ink data for the character, and the block 230 and the digital ink data can be used as the image.

The document generation unit 130 generates a document based on the character recognition result received from the character recognition unit 120 and the picture block 230 received from the document structure analysis unit 110. [ The document that can be generated by the document generation unit 130 may be a document in various file formats such as XML (eXtensible Markup Language), Microsoft Word, and PDF (Portable Document Format) as digital documents. The document generated by the document generation unit 130 may be stored in a memory, and may be used in conjunction with other systems or application programs at a later time.

The document digital restoration system and method according to an embodiment of the present invention can be used not only for recognition of print blocks and handwritten blocks but also for character recognition using digital ink data which is an online character recognition method, It is possible not only to increase the character recognition rate but also to minimize the inclusion of the image in the character recognition result, thereby improving the ease of editing.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module may reside in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a removable disk, a CD-ROM or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, which is capable of reading information from, and writing information to, the storage medium. Alternatively, the storage medium may be integral with the processor. The processor and the storage medium may reside within an application specific integrated circuit (ASIC). The ASIC may reside within the user terminal. Alternatively, the processor and the storage medium may reside as discrete components in a user terminal.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the present invention is not limited to the disclosed exemplary embodiments, but various changes and modifications may be made without departing from the spirit and scope of the invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the scope of the present invention but to limit the scope of the technical idea of the present invention. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

110: document structure analysis section
120: Character recognition unit
121: Printed character recognition unit
122: Handwritten character recognition unit
123: Digital ink data generation unit
124: Digital ink data character recognition unit
126:
127: Feature extraction unit
128:
129: Post-
130:
100: Document Digital Restoration System
200: Document image
210: print block
220: Handwriting Block
230: picture block

Claims (14)

A document structure analyzer for analyzing a document image and classifying a print block, a handwritten block, and a picture block from the document image;
A handwritten character recognizing unit for recognizing characters of the handwritten block; a handwritten character recognizing unit for recognizing a handwritten block which has failed in character recognition if the handwritten character recognizing unit fails in character recognition; A digital ink data generation unit for generating ink data, and a digital ink data character recognition unit for performing character recognition using the digital ink data, wherein the character recognition unit performs character recognition for each of the print block and the handwritten block part; And
And a document generation unit for generating a character recognition result from the character recognition unit and a document for the document image based on the picture block. The method according to claim 1,
Further comprising an input unit for receiving the document image,
Wherein the input unit is a scanner or a camera. The method according to claim 1,
The print character recognizing unit transmits a print block which failed in character recognition among the print block to the handwritten character recognizing unit or the digital ink data generating unit,
Wherein the handwritten character recognition unit further performs character recognition on the print block received from the print character recognition unit. The method of claim 3,
Wherein the handwritten character recognition unit transmits to the digital ink data generation unit a printed matter block in which a character recognition failure in the handwritten character block or a character block in the printed matter block received from the printed character recognition unit fails,
The digital ink data generation unit may further generate digital ink data for a print block received from the print character recognition unit, a handwritable block received from the handwritten character recognition unit, or a print block received from the handwritten character recognition unit , A digital restoration system for documents. The method according to claim 1,
The digital ink data character recognition unit, the handwritten character recognition unit, the digital ink data generation unit, and the digital ink data character recognition unit may include the print character recognition unit, the handwritten character recognition unit, the digital ink data generation unit, And a data character recognizing unit are sequentially driven in this order. The method according to claim 1,
Wherein the digital character data recognizing unit, the printed character recognizing unit, the handwritten character recognizing unit, the digital ink data generating unit, and the digital ink data character recognizing unit are driven in parallel. The method according to claim 1,
Wherein the printed character recognition unit and the handwritten character recognition unit are selectively activated. The method according to claim 1,
Wherein the print character recognition unit, the handwritten character recognition unit, and the digital ink data character recognition unit use a preprocessing process, a feature extraction process, a recognition process, and a post-process process for character recognition. The method according to claim 1,
Wherein the digital ink data character recognition unit transmits the digital ink data to the document generation unit when the character recognition fails. Analyzing the document image to classify a print block, a handwritten block, and a picture block from the document image;
Performing character recognition on the print block, performing character recognition on the handwritten block, generating digital ink data for a handwritten block failed in character recognition if the character recognition on the handwritten block fails Performing character recognition on each of the print block and the handwritten block, and performing character recognition using the digital ink data; And
And generating a document for the document image based on the character recognition result of the character recognition step and the picture block. 11. The method of claim 10,
Wherein the step of performing character recognition on the cursive block further comprises performing character recognition on a print block that fails to recognize the character when the character recognition on the print block fails. A method of digital restoration performed by a document. 11. The method of claim 10,
Wherein the step of generating digital ink data comprises the steps of generating digital ink data for a print block which fails to recognize a character when the character recognition for the print block fails, And generating digital ink data for the failed handwritten block. ≪ Desc / Clms Page number 19 > 11. The method of claim 10,
Performing character recognition on the print block, performing character recognition on the handwritten block, generating the digital ink data, and performing character recognition using the digital ink data are performed in parallel Wherein the digital restoration is performed with a document digital restoration system. Analyzing the document image to classify the print block, the handwritten block, and the picture block from the document image,
The digital ink data for the failed handwritten block is generated when the character recognition for the handwritten block is failed, the digital ink data for the failed handwritten block is generated, Performing character recognition using the digital ink data to perform character recognition on each of the print block and the handwritten block,
And a set of instructions for generating a document for the document image based on the character recognition result of performing the character recognition and the picture block.

Images