KR20090108487A - Method, apparatus and computer readable recording medium for preprocessing image - Google Patents

Abstract

PURPOSE: A method, an apparatus and a computer readable recording medium for preprocessing an image are provided to get a successful pre-processing result and improve the accuracy of text recognition. CONSTITUTION: A controller(110) controls the data flow among an image input unit(130), an image pre-processing unit(150), a text recognition unit(170) and a digital information generating unit(190). The image input unit(130) receives an image of a digitizing target object. The image pre-processing unit analyzes the logical structure included in an image inputted by the image input unit, and localizes at least some portion of the image. In addition, the image pre-processing unit determines a reference value of a pre-processing according to the logical structure and performs a pre-processing by applying the reference value to the localized image.

Description

METHOD, APPARATUS AND COMPUTER READABLE RECORDING MEDIUM FOR PREPROCESSING IMAGE}

The present invention relates to a method, an apparatus and a computer readable recording medium for image preprocessing. More particularly, the present invention relates to a method, an apparatus and a computer readable recording medium for analyzing a logical structure included in an image and performing image preprocessing based on the result.

In recent years, various technologies have been developed for processing text information included in an image by a digital device such as a computer. This process is commonly referred to as digitizing. Digitizing also refers to a series of processes for processing analog information contained in an image, document, voice, etc. into digital information.

In particular, in order to digitize an image, well-known Optical Character Recognition (OCR) technology is used (wherein optical character recognition is a hand included in an image generated by an optical device such as a scanner or a camera). Means a process of converting written or printed characters into characters that can be edited by a digital device such as a computer). Among these, character extraction technology, which is widely used in optical character recognition systems, is an important technique for determining the accuracy of character recognition, and research on this has been particularly actively conducted.

In order to extract and recognize a character from an image in an optical character recognition system or the like, first, the image must be preprocessed to extract an area including the character. Representative examples of the preprocessing techniques include binarization operations and edge detection. However, in general, the image is uneven characteristics of individual regions included in the image due to various reasons such as bad scanning state or bad typesetting state, so that if the preprocessing is adaptively performed according to the characteristics of each region, Results can be obtained, and research on this has been actively conducted.

Thus, an example of the technique which preprocesses an image in order to extract a character from an image is disclosed by Unexamined-Japanese-Patent No. 13-043312 (February 16, 2001, the invention name: document image processing apparatus).

In Japanese Patent Laid-Open No. 13-043312, after segmenting an input image and performing a first binarization operation on the entire divided image to find an area containing the character in the image, the character is A technique for extracting a character included in an image by performing a binarization operation on a second basis with different thresholds is disclosed.

And another example of a technique for preprocessing an image to extract a character from the image is disclosed in Japanese Patent Application Laid-Open No. 13-222683 (August 17, 2001. Publication, name of the invention: an image processing method, an image processing apparatus, a character recognition method, A character recognition device and a storage medium).

Japanese Patent Application Laid-Open No. Hei 13-222683 discloses that an input image is divided into a region in which no character exists and a region in which a character exists, and the threshold value of a binarization operation is adaptively determined for the region in which a character exists. By performing the operation, a technique for extracting characters contained in an image is disclosed.

In addition, an example of a technique for adaptively extracting and recognizing a character included in an image is disclosed in Japanese Patent Laid-Open No. 18-011966 (published on Jan. 12, 2006, title of the invention: a character recognition device, a character recognition program). have.

In Japanese Laid-Open Patent Publication No. 18-011966, character recognition is performed by decoding all or part of an input image at a first bit rate, and the accuracy of the character recognition is smaller than a predetermined value. In the following, a technique for performing character recognition again by decoding at a second bit rate higher than the first bit rate is disclosed.

However, according to the prior arts, including the techniques disclosed in the above publications, there is a problem that it is difficult to obtain satisfactory preprocessing results in some cases because the preprocessing of the image is performed regardless of the logical structure included in the image. (Here, the logical structure included in the image refers to a hierarchical structure composed of objects, such as spaces, tables, pictures, graphs, paragraphs, sentences, lines, words, characters, and phonemes, included in the image). For example, if the image is preprocessed without considering the logical structure included in the image, there is a problem that interference occurs between objects extracted by the preprocessing. In particular, this phenomenon often occurs for objects belonging to a relatively lower logical structure. In addition, if the preprocessing of the image is not successful, there is a problem that the accuracy of the final character recognition is also lowered.

The object of the present invention is to solve all the above-mentioned problems.

In addition, the present invention analyzes the logical structure included in the image prior to extracting the characters included in the image, and adaptively determine the reference value of the preprocessing according to the analyzed logical structure to pre-process the image, the successful preprocessing results The purpose is to improve the accuracy of character recognition.

In addition, an object of the present invention is to accurately analyze a logical structure included in an image using an adaptively preprocessed image.

Representative configuration of the present invention for achieving the above object is as follows.

According to an aspect of the present invention, an image preprocessing method, comprising: analyzing a logical structure included in an image, hierarchically regioning at least a portion of the image according to the analyzed logical structure, and analyzing the logical structure Adaptively determining a reference value for preprocessing the at least a portion of the imaged region according to the structure, and applying the determined reference value to the at least a portion of the imaged region .

According to another aspect of the present invention, an image preprocessing apparatus, comprising: a logical structure analyzing unit for analyzing a logical structure included in an image, a localization unit hierarchically localizing at least a portion of the image according to the analyzed logical structure; And a preprocessing operation unit adapted to adaptively determine a reference value for the preprocessing of the at least part of the image, according to the analyzed logical structure, and apply the determined reference value to the at least a portion of the image. An apparatus is provided.

According to still another aspect of the present invention, there is provided a digitizing apparatus, comprising: an image input unit for receiving an image, an image preprocessor for preprocessing the input image, a character recognition unit for recognizing a character from the preprocessed image, and the recognition And a digital information generating unit for generating digital information including the written characters, wherein the preprocessing unit comprises: a logical structure analyzing unit analyzing the logical structure included in the input image, the input image according to the analyzed logical structure And a localization unit hierarchically localizing at least a portion of the at least a portion, and adaptively determining a reference value for preprocessing the at least a portion of the image, and localizing the determined reference value according to the analyzed logical structure. It includes a preprocessing operation unit for applying to at least a portion of the image It is provided an apparatus.

According to the present invention, it is possible to adaptively perform image preprocessing according to the logical structure included in the image as described above, thereby minimizing the interference between objects extracted from the image, resulting in the recognition of the character from the image. The effect of improving accuracy is achieved.

In addition, according to the present invention, it is possible to analyze the logical structure of the input image using the adaptively pre-processed image.

DETAILED DESCRIPTION The following detailed description of the invention refers to the accompanying drawings that show, by way of illustration, specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It should be understood that the various embodiments of the present invention are different but need not be mutually exclusive. For example, certain shapes, structures, and characteristics described herein may be embodied in other embodiments without departing from the spirit and scope of the invention with respect to one embodiment. In addition, it is to be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the invention. The following detailed description, therefore, is not to be taken in a limiting sense, and the scope of the present invention, if properly described, is defined only by the appended claims, along with the full range of equivalents to which such claims are entitled. Like reference numerals in the drawings refer to the same or similar functions throughout the several aspects.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement the present invention.

[Preferred Embodiments of the Invention]

In the present specification, digitizing means a series of processes for processing analog information contained in an image, a document, a voice, and the like into digital information. In particular, the digitizing mentioned in the present specification not only generates a digital information including a character included in the image by recognizing the characters included in the image, but also considers the bibliographic information on the image or the logical structure included in the image, and the like. It should be understood in the broadest sense to include what is produced.

In addition, in this specification, the optical character recognition refers to converting handwritten characters or printed characters included in an image generated by an optical device such as a scanner or a camera into characters that can be edited by a digital device such as a computer. It means a series of processes. In particular, in the present specification, the optical character recognition should be understood as the broadest not only for recognizing characters included in an image, but also for recognizing objects such as tables and pictures.

In addition, in the present specification, the logical structure is a hierarchical structure composed of objects that an image may include, and the objects may be included in the image and may be optically distinguished, and may include blanks and tables included in the image. This can include pictures, graphs, paragraphs, sentences, lines, words, characters, and phonemes. In particular, it is to be understood that the logical structure referred to herein refers to the hierarchical structure between the objects, and that the objects can be zoned by position and size information within the image. However, logical structures do not have to be hierarchical, and one object does not necessarily have to be included in another.

Configuration of the entire system

1 is a schematic configuration of an entire system for performing digitizing by adaptively preprocessing an image according to a logical structure included therein and extracting and recognizing characters from the preprocessed image according to an embodiment of the present invention. It is also.

As shown in FIG. 1, the entire system according to an embodiment of the present invention includes a digitizing apparatus 100 for digitizing an image including a character, an image database in which an image to be digitized is stored ( 200) and a digital information database 300 in which digital information generated as a result of digitizing is stored.

First, the digitizing apparatus 100 according to an embodiment of the present invention may receive an image and extract a character from the input image to generate digital information. Specifically, the digitizing device 100 may be a digital device (not shown) itself having a computing power by mounting a processor and having a memory means such as an industrial / personal computer, a workstation, a PDA, a mobile phone, or the like. It may also be related to a program module included in the. Such program modules may be included in digital devices in the form of operating systems, application modules and other program modules, and may be physically stored on various known storage devices. In addition, the program module may be stored in a remote storage device that can communicate with the digital device. On the other hand, such program modules include, but are not limited to, routines, subroutines, programs, objects, components, data structures, etc. that perform particular tasks or execute particular data types, which will be described later, in accordance with the present invention.

As described above, the image database 200 and the digital information database 300 respectively store the image to be digitized and digital information generated as a result of digitizing.

The image according to the present invention is an object of digitizing, and refers generically to digital information further including not only the image itself but also information such as the source, bibliographic information, creation date, and the like. An image is generally generated by scanning or photographing a hand-written image or a printed image by using an optical device (not shown) such as a scanner or a camera. However, the image is not necessarily limited thereto. Any digital information may correspond to the image in the present invention as long as it can be an object. For example, an image generated by an input device (not shown) such as a mouse, pen mouse, touch pad, touch panel, tablet, or the like may also be included in the image according to the present invention. In addition, the image does not necessarily need to include content written in the form of a document, and the image may be a color image, a gray-scale image, or an image that may be processed by a digital device such as a computer. All may be included.

In addition, the digital information according to the present invention collectively refers to information generated by digitizing the image by the digitizing apparatus 100 of the present invention, and includes bibliographic information about the image, a logical structure included in the image, and an image. It should be understood to encompass the recognition information of the characters contained.

Digitizing  Configuration of the device

Hereinafter, the internal structure of the digitizing apparatus 100 and the function of each component will be described.

Referring to FIG. 1, the digitizing apparatus 100 according to an embodiment of the present invention includes a control unit 110 for controlling each component in the digitizing apparatus 100, an image input unit 130 for receiving an image, and an input image. Image preprocessing unit 150 for performing the preprocessing, character recognition unit 170 for recognizing the characters included in the image using the pre-processed image, bibliographic information on the image, logical structure information included in the image and It can be seen that it can be configured as a digital information generating unit 190 for generating digital information according to a predetermined format using the recognition information of the characters included in the image.

According to an embodiment of the present invention, the control unit 110, the image input unit 130, the image preprocessor 150, the character recognition unit 170 and the digital information generating unit 190 at least some of them are program modules Can be. Such program modules may be included in the digitizing device 100 in the form of operating systems, application modules, and other program modules, and may be physically stored on various known storage devices. In addition, such program modules may be stored in a remote storage device that can communicate with the digitizing device 100. On the other hand, such program modules include, but are not limited to, routines, subroutines, programs, objects, components, data structures, etc. that perform particular tasks or execute particular abstract data types, described below, in accordance with the present invention.

First, the controller 110 controls the flow of data between the image input unit 130, the image preprocessor 150, the character recognizer 170, and the digital information generator 190. That is, the control unit 100 according to an embodiment of the present invention controls the flow of data between the components of the digitizing apparatus 100, so that the image input unit 130, the image preprocessor 150, the character recognition unit 170 And the digital information generating unit 190 each perform its own function.

According to an embodiment of the present invention, the image input unit 130 performs a function of receiving an image that is a target of digitizing. The type of image to be digitized does not necessarily need to be limited, but preferably, a file of an image or a set of images including characters may be considered.

The image preprocessor 150 analyzes the logical structure included in the image input by the image input unit 130 and localizes at least a portion of the image according to an embodiment of the present invention. In accordance with the present invention, the reference value of the preprocessing may be adaptively determined, and the preprocessing may be performed by applying the same to the localized image. The image preprocessor 150 according to an embodiment of the present invention will be described in detail later with reference to FIGS. 2 and 3.

On the other hand, according to an embodiment of the present invention, the character recognition unit 170 recognizes the characters contained in the image pre-processed by a known image analysis algorithm. In this regard, those skilled in the art to which the present invention belongs may refer to the character recognition technology disclosed in the Korean Information Science Society, Vol. 34, No. 1 (C), which is a well-known technical document. have. Of course, with respect to the image analysis algorithm, it is possible to refer to various documents in addition to the above documents.

According to an embodiment of the present invention, the digital information generating unit 190 uses digital information according to a predetermined format by using bibliographic information on an image, a logical structure included in the image, and recognition information of characters included in the image. Performs the function of generating it. For example, the digital information according to an embodiment of the present invention may be implemented as a document file such as a word file, or a web page expressed in a markup language such as HTML (HyperText Markup Language) or XML (eXtensible Markup Language). It may be implemented as.

Image Preprocessor

Hereinafter, the internal structure of the image preprocessor 150 and the function of each component will be described.

2 is a diagram illustrating an internal configuration of the image preprocessor 150 according to an embodiment of the present invention in detail. Referring to FIG. 2, it can be seen that the image preprocessor 150 according to an embodiment of the present invention may include a preprocessing operator 152, a logical structure analyzer 154, a localizer 156, and the like. have.

First, before processing the character, the preprocessing operation unit 152 according to an embodiment of the present invention accurately extracts an image region including the character so that the character included in the image is clearly distinguished from a portion other than the character. It performs the function so that it can be processed. The preprocessing operation unit 152 may employ various preprocessing techniques for the above-described processing. As an example of such a preprocessing technique, a binarization operation may be assumed. Here, the binarization operation means a process of specifying a pixel value of 1 when the pixel value is greater than or equal to a threshold value for an individual pixel constituting the image, and specifying a pixel value of 0 otherwise. In general, since a portion corresponding to a character is relatively dark in the image, and thus has a relatively low pixel value, a binarization operation can usually emphasize the character portion in the image. Of course, not only the binarization operation may be employed as such a preprocessing technique, but also a technique capable of clearly distinguishing a character part included in an image, and may be adopted as the preprocessing technique of the present invention.

According to an embodiment of the present invention, in performing the preprocessing on the image, the preprocessing operation unit 152 may adaptively apply the reference value for the preprocessing according to the property of the specific region of the image. Specifically, in relation to the reference value for the preprocessing, when the binarization operation is employed as the preprocessing technique according to an embodiment of the present invention, the preprocessing may be performed on the image while adaptively changing the threshold value of the binarization. .

More specifically, in relation to an attribute of a specific region of the image, according to an embodiment of the present invention, according to the logical structure corresponding to the specific region of the image (analyzed by the logical region analyzing unit 154 described later) The threshold for the binarization operation may be adaptively applied. Preferably, the threshold value of the binarization operation may be applied to the region corresponding to the logical structure that is relatively higher, and the threshold value of the binarization operation may be adjusted lower to the region corresponding to the logical structure that is relatively lower. Can be applied.

According to the present invention, the logical region analyzer 154 analyzes the hierarchical structure of the objects included in the image, that is, the logical structure included in the image. The logical region analyzer 154 according to an embodiment of the present invention may analyze a logical structure included in an image by a known algorithm. In this regard, those skilled in the art to which the present invention pertains are known. See, for example, the article "Computer Vision and Pattern Recognition, 1994" by Robert M. Haralick, "Document Image Understanding: Geometric and Logical Layout," and the like. Of course, those who intend to practice the present invention is not limited thereto, and may employ various known techniques to implement the present invention.

According to an embodiment of the present invention, the localization unit 156 performs a function of localizing an image according to the logical structure of the image analyzed by the logical structure analyzer 154. According to an embodiment of the present invention, the localization unit 156 merges and localizes portions corresponding to the logical structure of the same layer among the images, and the logical structure of the image is hierarchical and belongs to the logical structure of the image. Since the objects may form an inclusion relationship with each other, the regions generated by the localization unit 156 may have portions overlapping each other. In this regard, as a technique of localizing a specific portion of an image, a connected component labeling technique or the like may be considered. However, of course, not only connection element labeling techniques are available to implement the present invention, but various modifications may be applied to implement the present invention.

Hereinafter, a detailed operation principle and preprocessing performance of the image preprocessor 150 will be described in detail with reference to the accompanying drawings.

3A is a diagram sequentially illustrating an operation of the image preprocessor 150 according to an embodiment of the present invention, and FIG. 3B is a diagram illustrating a result of preprocessing an image performed by the image preprocessor 150. Drawing.

Referring to FIG. 3A, the image preprocessor 150 primarily performs preprocessing on the input image (S301), and analyzes an upper logical structure of the first preprocessed image (S303). Typically, the objects corresponding to this higher logical structure may correspond to a relatively wide area in the image. For example, paragraphs, sentences, lines, etc. may be the objects corresponding to the higher logical structure in the present invention. Can be. Then, the image is regioned according to the analyzed logical structure (S305). Accordingly, an area corresponding to a specific logical structure in the image may be distinguished from other areas.

Next, by preliminarily determining and applying the reference value of the preprocessing according to the logical structure included in the areaized image, the preprocessing is performed second (S307). Preferably, such secondary preprocessing may be performed on the input original image, rather than on the image that is output as a result of the primary preprocessing. Of course, less preferably, secondary preprocessing may be performed on the image that is output as a result of the primary preprocessing. In addition, since the secondary preprocessing may be performed only on the image region belonging to a specific logical structure, more detailed and accurate preprocessing results may be obtained as compared with the case of the first preprocessing. Subsequently, the lower logical structure included in the second preprocessed image may be analyzed (S309). In general, such a lower logical structure corresponds to a relatively narrow area in the image. For example, words, letters, phonemes, etc. may be objects belonging to the lower logical structure of the present invention.

The exemplary image preprocessing process performed by the image preprocessor 150 is not necessarily limited to that shown in FIG. 3A. For example, the preprocessing may be repeatedly performed in order from the highest logical structure to the lowest logical structure among the logical structures included in the image. Here, the object corresponding to the highest logical structure may be the input image itself, or the object corresponding to the lowest logical structure may be a letter or a phoneme.

Referring to FIG. 3B, a result of the first preprocessing of the image and a result of the second preprocessing may be compared. For example, the results of performing the preprocessing on the word "soliciting" will be specifically compared. According to FIG. 3B, as a result of performing the first preprocessing fixedly applying the reference value of the preprocessing without considering the logical structure of the image, in the word "soliciting", all the consonants and vowels of the letter "blue" are connected, and the letter "bottom" The consonants of "are connected, and the letters" ha "and the letter" a "are connected to each other. As described above, when letters or phonemes which are constituents of words are connected to each other and extracted, it is difficult to separate and recognize them. As a result, the accuracy of character recognition is lowered. On the other hand, in the case of performing the second preprocessing by adaptively determining and applying the reference value of the preprocessing according to the logical structure included in the image, it can be seen that all the components of the word "listening" are separated from each other without being connected to each other. have. As described above, when the components of the words are separated from each other and extracted, the components of the words can be recognized separately, resulting in an increase in the accuracy of character recognition.

As described above, the image preprocessor 150 according to an embodiment of the present invention can adaptively preprocess an image by adaptively determining and applying a reference value of preprocessing according to the logical structure of the image.

Application Examples of the Invention

Hereinafter, according to an embodiment of the present invention, the digitizing apparatus 100 receives an image, performs preprocessing, recognizes characters included in the image, and generates an example of digital information corresponding to the image. Let's look at it.

4 is a flowchart illustrating a sequential operation of the digitizing apparatus 100 according to an embodiment of the present invention, and FIG. 5A illustrates a result of performing a binarization operation on an image according to an embodiment of the present invention. 5B is a diagram illustrating a result of performing segmentation by analyzing a logical structure of an image according to an embodiment of the present invention, and FIG. 5C is a diagram illustrating digitizing according to an embodiment of the present invention. It is a figure which shows the resulting digital information as an example.

For example, assume that the digitizing device 100 receives an image from the image database 200 (S400). In this case, the image may be input by the image input unit 130. At this time, according to an embodiment of the present invention, not only the image but also information such as the source, the bibliographic information, the creation date, and the like may be input together.

Next, according to an embodiment of the present invention, the preprocessing operation unit 152 of the image preprocessor 150 may preprocess the image by applying a preset reference value of the preprocess (S402). According to a preferred embodiment of the present invention, as a technique for preprocessing an image, a binarization operation may be adopted, and a threshold value of the binarization operation may be adopted as a reference value of the preprocessing.

Next, according to an embodiment of the present invention, the logical structure analyzing unit 154 of the image preprocessor 150 may analyze the logical structure of the image formed by the objects extracted by the preprocessing of the previous step ( S404). According to a preferred embodiment of the present invention, the logical structure may be a hierarchical structure formed by objects such as spaces, tables, figures, graphs, paragraphs, sentences, lines, words, characters, and phonemes. The logical structure analyzing unit 154 may analyze only the logical structure that is relatively higher, and may analyze only the highest logical structure as needed.

Next, according to an embodiment of the present invention, the localization unit 156 of the image preprocessor 150 merges portions corresponding to the logical structure of the same layer among the preprocessed images according to the analyzed logical structure. It can be zoned (S406). According to an exemplary embodiment of the present invention, the regions generated by the localization unit 156 may have portions overlapping each other.

Next, according to an embodiment of the present invention, the preprocessing operation unit 152 of the image preprocessing unit 150 iteratively by adaptively determining and applying the reference value of the preprocessing according to the logical structure corresponding to the areaized portion. Phosphorus pretreatment can be performed. That is, the preprocessing may be repeatedly performed on the area of the input image (S408). According to a preferred embodiment of the present invention, the repetitive preprocessing as described above does not have to be performed on the image outputted as a result of the preprocessing in the previous step, but on the input original image that has not yet been preprocessed. Can be performed. In addition, according to an exemplary embodiment of the present invention, the repetitive preprocessing as described above may be performed on the entire input image, or may be performed only on the region of the input image. In addition, according to an exemplary embodiment of the present invention, the repetitive preprocessing as described above may be performed by applying a threshold lower than the threshold value of the binarization operation performed in the previous step, when the preprocessing is a binarization operation. It may be performed by applying a threshold higher than the threshold of the binarization operation performed in the previous step.

Next, according to an embodiment of the present invention, when the preprocessing performed in the previous step is performed for the area corresponding to the lowest logical structure, the character recognition step S410 may be performed, and the lowest logical structure If it is not performed for the area corresponding to the step of analyzing the logical structure (S404) may be further performed.

As such, when the image preprocessor 150 performs preprocessing on the input image, the character recognition unit 170 may recognize a character included in the preprocessed image, and the digital information generation unit 190 may input the image. Digital information can be generated using information such as the source, bibliographic information, creation date, and the like of character recognition information.

Embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed by various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the computer-readable recording medium may be those specially designed and configured for the present invention, or may be known and available to those skilled in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tape, optical recording media such as CD-ROMs, DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules to perform the process according to the invention, and vice versa.

Although the present invention has been described by specific embodiments such as specific components and the like, but the embodiments and the drawings are provided to assist in a more general understanding of the present invention, the present invention is not limited to the above embodiments. For those skilled in the art, various modifications and variations can be made from these descriptions.

Accordingly, the spirit of the present invention should not be limited to the above-described embodiments, and all of the equivalents or equivalents of the claims, as well as the appended claims, fall within the scope of the spirit of the present invention. I will say.

1 is a schematic configuration of an entire system for performing digitizing by adaptively preprocessing an image according to a logical structure included therein, extracting and recognizing characters from the preprocessed image, according to an embodiment of the present invention. It is also.

2 is a diagram illustrating in detail the internal configuration of the image preprocessor according to an embodiment of the present invention.

3A is a diagram sequentially illustrating an operation of an image preprocessor according to an exemplary embodiment of the present invention, and FIG. 3B is a diagram illustrating a result of performing preprocessing on an image by the image preprocessor.

4 is a flowchart illustrating a sequential operation of the digitizing apparatus according to an embodiment of the present invention.

5A is a diagram exemplarily illustrating a result of performing a binarization operation on an image according to an embodiment of the present invention, and FIG. 5B is an area analysis by analyzing a logical structure of an image according to an embodiment of the present invention. FIG. 5C is a diagram illustrating one result, and FIG. 5C is a diagram illustrating digital information generated as a result of digitizing according to an embodiment of the present invention.

<Brief description of the major reference numerals>

110: control unit

130: image input unit

150: image preprocessor

152: preprocessing unit

154: logical structure analysis unit

156: zone

170: character recognition unit

190: digital information generating unit

Claims (11)

Image preprocessing method, Analyzing the logical structure contained in the image, Hierarchically regioning at least a portion of the image according to the analyzed logical structure, Adaptively determining a reference value for preprocessing the at least a portion of the image in accordance with the analyzed logical structure, and Applying the determined reference value to at least some of the imaged regions How to include. The method of claim 1, The logical structure is a hierarchical structure of the objects included in the image. The method of claim 2, Each of said objects being one of a space, a table, a picture, a graph, a paragraph, a sentence, a line, a word, a character and a phoneme. The method of claim 3, Said image preprocessing is a binarization operation. The method of claim 4, wherein The reference value is a threshold of a binarization operation. The method of claim 5, And determining the reference value such that the threshold value for the upper logical structure included in the image is not smaller than the threshold value for the lower logical structure included in the image. The method of claim 1, And determining the reference value and applying the reference value are sequentially performed from the highest logical structure to the lowest logical structure included in the image. The method of claim 1, Performing preprocessing on the image according to the applied reference value How to include more. A computer-readable recording medium for recording a computer program for executing the method according to any one of claims 1 to 8. Image preprocessing device, Logical structure analysis unit for analyzing the logical structure included in the image, A localization unit hierarchically localizing at least a portion of the image according to the analyzed logical structure, and According to the analyzed logical structure, a preprocessing operation unit that adaptively determines a reference value for preprocessing the at least a portion of the image, and applies the determined reference value to the at least a portion of the image, Device comprising a. As a digitizing device, Image input unit for receiving an image, An image preprocessing unit performing preprocessing on the input image; A character recognition unit recognizing characters from the preprocessed image, and Digital information generation unit for generating digital information including the recognized character Including but not limited to: The preprocessing unit, A logical structure analyzer for analyzing a logical structure included in the input image; A localization unit hierarchically localizing at least a portion of the input image according to the analyzed logical structure, and According to the analyzed logical structure, a preprocessing operation unit that adaptively determines a reference value for preprocessing the at least a portion of the image, and applies the determined reference value to the at least a portion of the image, Device comprising a.

Images