KR20090029430A - Method and apparatus of digitizing electrocardiogram papers to interwork with electronic medical record system - Google Patents

Abstract

A binary coded method and an apparatus of the electrocardiogram recording paper, capable of connecting to an electronic medical record system are provided to data the text information and the record line information of an electrocardiogram recording paper in the binary. A database(104) stores the information of an electrocardiogram recording paper. If the image information according to the scan of the electrocardiogram recording paper is provided, a controller(100) performs the preprocess of rotation or size control etc. by comparing with the grid pattern of the standard electrocardiogram recording paper after extracting the grid for the image-information process from the image information, and detects the base line by removing the grid about the preprocessed image information. The controller divides the image information from which the grid is removed into the character information field and the record line information field, and extracts the character information is extracted the character information according to the OCR(Optical Character Reader) technology from the image information.

Description

Method and apparatus for digitizing electrocardiogram papers interoperable with electronic medical record system {Method and apparatus of digitizing electrocardiogram papers to interwork with Electronic Medical Record system}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electronic medical recording technology, and more particularly, obtains image information on an electrocardiogram recording paper through a scanner, etc., and obtains text information and recording line information (graph) of the electrocardiogram recording paper from the image information. A method and apparatus for extracting and binarizing and processing into a form suitable for an electronic medical recording environment.

In general, an ECG recording paper records text information such as patient information and measurement information, and recording line information (graph) that graphically displays a signal output through an ECG lead. In particular, the recording line information is displayed on a grid of regular intervals and its size and shape are very important factors for diagnosis.

The current method of digitizing ECG recording paper is to scan ECG recording paper as an image and store it as an image file, which has the same effect as seeing ECG recording paper directly by accurately storing the information on ECG recording paper itself. However, when storing a high resolution image file for accurate diagnosis and discrimination, the file size becomes very large.

Currently, the digital storage method of one-dimensional biosignals such as electrocardiograms is being enacted by the international standard ISO / DTS 11073-90201, and data storage in the form of character information rather than image information is required. In the case of video information, data is stored for empty spaces that do not have valid information. As a result, the loss of data cannot be reconstructed in the undamaged parts due to the disadvantages of storage capacity and damage of unnecessary parts. Because there was a problem. On the other hand, in case of storing as character information, even if an error of a specific bit occurs, diagnosis and discrimination using the remaining data except the data can be performed, and by storing only necessary data, the capacity of the storage file can be minimized.

Since the international standard is limited to the case where a digital signal is obtained, a technique of binary digitizing and storing both text information and graph of an electrocardiogram recording paper output in an analog form is required.

Current domestic applications and registered patents related to the technology are as follows.

-Patent Registration No. 10-0538577 Computerized Standardization Method of Medical Information

-Patent Registration No. 10-0544512 Device-oriented electronic medical record data composition device and method for inquiring thereof

-Patent Application No. 10-2001-0049217 Telemedicine system and method using the network

The “electronic medical record providing method” (Registration No. 10-0600863) acquired by Park Yong-nam of the patent relates to a method and system for building an electronic medical medical record database using a public key infrastructure encryption algorithm of 128 bits or more, or a scanner of the system. There is no mention of the method of constructing the electronic medical medical record of ECG recording papers, and the electronic medical record is also recorded in “Patient-centered electronic medical record data composition device and method of obtaining it” (Reg. No. 10-0544512) acquired by MDWARE Co., Ltd. Reference is made to electronic mandatory records, including the system's constituent devices and inspection records for organizing data, but there is no mention of electronic data formation of electrocardiograph recordings. In addition, even in the case of the "telephone system and the method using the network" (Application No. 10-2001-0049217) filed by Hyundai Medin Co., Ltd. of the patent there is no mention of the electronic obligation recording of the existing ECG recording paper.

As described above, there is no patent regarding a method for storing an ECG recording paper as binary text data.

In addition, foreign patents related to the above are as follows.

US 6,792,574 B1 COMPUTER-BASED PATIENT RECORDING SYSTEM

US 5,832,450 ELECTRONIC MEDICAL RECORD USING TEXT DATABASE

-US 7,133,937 INPUT DEVICES FOR ENTERING DATA INTO AN ELECTRONIC MEDICAL RECORD

-US 6,573,887 B1 COMBINED WRITING INSTRUMENT AND DIGITAL DOCUMENTOR

US 6,347,329 B1 ELECTRONIC MEDICAL RECORDS SYSTEM

US Patent No. 6,792,574, obtained by Sanyo Electric Co., Ltd., relates to the structure of various menus and programs for computer storage and display of patient information, and US Patent 5,832,450, obtained by Scott & White Memorial Hospital, is text-based. Of US Patent No. 7,133,937, issued by GE Medical Systems Information Technologies, and US Patent No. 6,573,887, filed by O'Donnell, Jr., provide a display device that includes a writing instrument type input device and a touch pad. With respect to the electronic medical record preparation device and system used, the above-mentioned application and registered patents do not have any application of the electronic medical record of the electrocardiograph recording paper. US Patent 6,347,329, obtained by Macneal Memorial Hospital Assoc., Is a patent on the field inspection device and electronic medical record storage of various patient information. There is no mention of ECG recording paper, but the data acquired through the scanner is used as image information data. Since the same as the above-mentioned method of processing ECG recording paper in use, it is the same as the above-mentioned foreign applications and registered patents, the method for storing full binary text data of ECG recording paper presented in the present invention is not shown. .

Accordingly, the present invention has been proposed to enable the storage, management, and various signal processing of data acquired from a digital electrocardiograph and data of an existing electrocardiogram recording paper in the same form while solving the conventional problems as described above. It is an object of the present invention to provide a method and apparatus for binarizing an electrocardiogram recording paper which can be linked to an electronic medical recording system for binaryizing text and record line data.

The binarization method of an electrocardiogram recording paper of the present invention for achieving the above object comprises the steps of: pre-processing the electrocardiogram recording paper image information; Removing a grid from the preprocessed image information; Extracting a baseline from the image information from which the grid is removed; Extracting a recording line information (graph) region and a text information (text) region from the image information from which the grid is removed; Recognizing text information from the video information of the text information area; Binarizing the image information of the recording line information area or extracting the minimum area based on the base line; And selectively storing the character information and the binary record line information or the character information and the minimum area in a database.

The data of the electrocardiograph recording paper processed by the method of the present invention as described above becomes a one-dimensional output signal with respect to time as the output of the digital electrocardiograph. The present methods are not able to apply various signal processing methods using the characteristic of ECG signal which is one-dimensional signal by storing the data of output paper as two-dimensional image signal. In addition, it is not possible to discriminate and to process various statistics and signals from the electronic medical records in the future. However, when the graph of ECG recording paper is stored as a 1D signal using the method of the present invention, it is not only applicable to various signal processing and automatic diagnosis algorithms, but also based on the contents stored in the electronic medical record, and clinical treatment appropriate for each patient's condition. Even the Clinical Pathway (CP) can be provided.

As described above, according to the present invention, the entire ECG recording paper may be binarized or stored, or only the minimum graph area may be extracted and stored. Therefore, the contents of ECG recording paper can be divided into character information and recording line information, and the character recognition, baseline detection, and binarization can be extracted and stored, respectively, which can significantly reduce the capacity of the storage file for long-term storage of patient records. It is a technical basis that it is easy to restore the information about the undamaged part even when the stored file is damaged due to the base storage. In addition, since the relationship between the information is not complex two-dimensional information such as image information, it is a binarization information that is arranged over time, so that it can be applied to automatic extraction from a database and various automatic diagnostic techniques, and various statistics in an electronic medical record environment. Application to data and automatic processing of personal health records is possible.

In addition, the present invention can be easily converted to the storage format to meet the standards for sharing personal health information because the ECG recording paper can be stored as binary information, so it is easy to apply to the storage method suitable for various standards to be converted data There is an effect that can reduce the overall cost and time effort.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above-described embodiments, which can be variously modified and modified by those skilled in the art to which the present invention pertains. Modifications are possible.

Therefore, the spirit of the present invention should be grasped only by the claims described below, and all equivalent or equivalent modifications thereof will belong to the scope of the present invention.

A method and apparatus for binarizing an electrocardiogram recording paper interoperable with an electronic medical recording system according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of an electrocardiogram recording paper binarization apparatus according to a preferred embodiment of the present invention.

The electrocardiogram recording binarization apparatus is largely composed of a control device 100, a scanner 102, a database 104, a display device 106, a communication module 108, a memory unit 110, and a user interface device 112. .

When the image information on the ECG recording paper is provided from the scanner 102, the control device 100 extracts the text information and the recording line information (graph) of the ECG recording paper and converts the data into binary data. Store in

The scanner 102 scans an electrocardiogram recording paper under the control of the control device 100 to generate image information, and provides the image information to the control device 100. In particular, the scanner 102 may be connected through a network. In this case, the ECG recording paper data may be generated by a method of scanning and transmitting the ECG recording paper from a remote location. In addition, an imaging device may be provided instead of the scanner 102.

The database 104 stores the electrocardiogram recording paper information generated by binarizing the electrocardiogram recording paper under the control of the control device 100. The ECG recording sheet information is composed of ECG recording sheet identification information, character information, and recording line information.

The display device 106 displays electrocardiogram recording sheet information under the control of the control device 100 to guide the user.

The communication module 108 transmits the electrocardiogram recording paper information stored in the database 104 to an external server or an external terminal through a network such as the Internet under the control of the control device 100.

The memory unit 110 stores a variety of information including the processing program of the control device 100, and stores the standard grid image information for the present invention.

The user interface device 112 receives various commands from a user with a keyboard or a mouse and provides the same to the control device 100.

An electrocardiogram recording paper binarization method according to a preferred embodiment of the present invention applicable to an electrocardiogram recording paper binarization apparatus configured as described above will be described with reference to FIG.

When the ECG recording sheet processing command is received through the user interface device 112 (step 20), the controller 100 drives the scanner 102 to scan the ECG recording sheet (step 22).

When the image information according to the scan of the ECG recording paper is provided, the controller 100 extracts a grid forming a grid for the image information processing from the image information and rotates or adjusts the size of the ECG recording paper in comparison with the grid pattern of the standard ECG recording paper. In addition to performing preprocessing, the grid is removed from the preprocessed image information, and a baseline is detected (step 24). Thereafter, the control unit 100 divides the image information from which the grid is removed into a text information area and a recording line information graph area (step 26), and reads an optical character from the image information of the text information area (OCR). character information is extracted according to an optical charactoer reader technology (step 28).

Thereafter, the controller 100 checks the processing mode of the recording line information (step 30). The processing mode of the recording line information includes a mode for extracting the minimum region among the areas where the recording line information is located and a mode for binarizing the area where the recording line information is located. The mode is set according to a user's selection.

If the processing mode of the recording line information is the minimum area extraction mode, the control unit 100 extracts the minimum area for the recording line information area (step 32), and extracts the image information of the extracted minimum area and the extracted text information area. Corresponding character information is stored in the database 104 (step 34).

If the processing mode of the recording line information is the binarization mode, the controller 100 extracts the guide information from the image information of the recording line information area, divides the guide information area, and displays the image information of the guide information area. Implement binarization (steps 36-38). When the binarization is completed, the control unit 100 associates the binarization information of the image information of the region for each guide information with the character information of the extracted character information region and stores it in the database 104 (step 40).

3A illustrates the ECG recording sheet image information 200. The ECG recording sheet outputs all of the induction signals of the standard 12 leads on one screen, and the induction signal 210 is located above the ECG graph. Signal 220 is recorded for 3 seconds. The order of the induction on the ECG recording sheet is specified as a standard. The base of the recording paper also includes a grid 230 in the form of a grid already printed on the recording paper. Accordingly, the area of each lead in the image information of the ECG recording paper is divided based on the vertical dividing line or the lead information, the vertical dividing line, and the lead information positioned above the base line.

FIG. 3B illustrates an example in which character information 240 such as patient and measurement information, which is not displayed in FIG. 3A, is included. Referring to FIG. 3B, the character information is mostly in a portion without a grid scale, but the measurement information portion is located in a grid 230 in a grid scale form at the bottom of the paper. Since the character information 240 present in the grid scale part is also important information related to measurement, it must be stored as text through optical character processing.

4 is a detailed view of image information of a grid 230 portion of the ECG recording sheet. The grid 230 is a sequence of regular squares having vertical and horizontal square structures. The grid 230 is composed of a thick reference line 250 and a thin grid line 260. In the standard ECG output paper, the small scale is 1 mm in both horizontal and vertical, and the thick scale is 5 mm. Since the grid spacing is very important information informing the size and shape of the ECG signal actually output, it is necessary to reflect the data and make data. However, it is advantageous to remove the grid because errors may be caused by the grid while actually performing optical character reading or converting the ECG graph to binary information. In addition, since the grid spacing increases or decreases in the scanned image information as the image information is enlarged and reduced as described above, it is necessary to standardize it. In the present invention, a method for removing the grid while standardizing the grid size is presented as follows.

5A is a detailed flowchart of image information preprocessing and grid removing process according to a preferred embodiment of the present invention. FIGS. 5B, 5D, and 5E show regions of image information in which recording lines are located, and FIG. FIG. 5G illustrates grid image information, and FIG. 5G illustrates an example of image information from which a grid is removed.

5A to 5G, the image information preprocessing and grid removing process will be described in detail.

First, the controller 100 checks whether the grid of the electrocardiogram recording paper image information and the standard grid image information 300 match (ST1).

If the grid of ECG image paper image information and the standard grid image information 300 do not match, the controller 100 detects the degree of rotation of the document caused in the process of scanning the ECG recording paper (ST2). That is, the controller 100 detects the degree of rotation of the ECG image paper image information by comparing horizontal and vertical lines with respect to the grid of ECG image paper image information and the grid of standard grid image information 300, and according to the degree of rotation. The image information is reversely rotated to correct rotation of a document caused during the scanning process of the ECG recording sheet (ST3).

After correcting the degree of rotation of the document, the controller 100 returns to (ST1) to check whether the grid of ECG image paper image information and the standard grid image information 300 match, and if the two grids still do not match The size of the image information is adjusted (ST4). That is, the controller 100 adjusts the size of the image information so that the size of the grid of the standard grid image information 300 and the size of the grid of the ECG recording paper image information are the same.

When the grid of the ECG recording sheet image information matches with the standard grid through the above process, the controller 100 removes the grid from the ECG recording sheet image information. The removal of the grid may include obtaining a difference of the standard grid image information from the image information of the ECG recording paper, or deleting pixels corresponding to the color of the grid as the color of the grid is distinguished from the character information or the recording line information. Can be.

5B to 5G illustrate image information of an ECG recording paper to help understand each step of FIG. 5A. The rotation, size normalization, and grid removal of image information for a portion 270 of the ECG recording paper image information are illustrated. The process is illustrated.

The electrocardiogram recording image information 270 is compared with the standard grid grid image information 310 and detects and corrects the rotation degree through the horizontal reference line 311 and the vertical reference line 312 to correct the rotation degree image information 320. Get The image information 320 whose rotation degree is corrected is enlarged and reduced while maintaining the aspect ratio, thereby obtaining the image information 330 matching the grid size of the standard grid image information 310. If the difference between the image information 330 and the standard grid image information 310 is obtained, the image information 340 from which the background grid is removed may be obtained. In addition, when the background grid 230 has a color distinguished from other recordings on the ECG recording paper, the image information 340 from which the background grid is removed may be obtained by selectively erasing the color itself.

The process of binarizing the image information of the recording line information area will now be described.

FIG. 6 illustrates a histogram in a horizontal (time axis) direction of image information from which a grid of a background is removed.

The histogram may be obtained by adding brightness information of each pixel in a horizontal (time axis) direction of the image information 340 from which the grid is removed.

Four peak values appear in the histogram, and the positions of the peaks become the baseline of each ECG recording line. Here, a predetermined threshold is given to the peak value, thereby minimizing that the baseline is set incorrectly by an error due to scanning or imaging.

In the electrocardiograph, since most of the energy is distributed in the baseline, the baseline is extracted and the number of pixels from the baseline to the pixel where the recording line is located is detected. The number of pixels from the baseline to the pixel where the recording line is located becomes binary data of the baseline.

Next, a process of extracting only the minimum area of the image information of the recording line information area will be described.

The base line is detected from the image information, the number of pixels of the recording line farthest from the recording line information is detected as the base line, and the vertical range corresponding to the number of pixels is set to the minimum area.

Next, a process of discriminating the character information area and the recording line information area from the image information will be described.

The present invention distinguishes between a text information area and a recording line information area based on the complexity thereof by using an image information window having a specific pixel size. That is, in the case of the recording line information, since the one-dimensional change curve with respect to time, the shapes appearing in the neighboring pixels are represented by only one single line, and in the case of the character information, several lines appear at once. Do. Accordingly, the present invention discriminates the character information area and the recording line information area based on the complexity of the image information.

In addition, a text information area and a recording line information area may be determined using the base line and the image information window at the same time. That is, by extracting the baselines of the four ECG graphs by the baseline extraction method and using the complexity determination method for the areas outside the four baselines, character information other than the recording line information can be easily and quickly determined. In addition to the above examples, the above methods may be applied in part or in whole, or may be used alone or in combination to produce various performances.

FIG. 7 illustrates an example in which the minimum range 710 is extracted from the recording line information area in the segmented graph area image information 550, and an example 910 of character information obtained through data conversion by binarizing the recording line information. It is shown.

As shown in FIG. 7, the minimum region extraction from the image information 550 from which the character information is removed can be performed by using a dividing line between each lead. Since the dividing line between each lead is perpendicular to the baseline of the ECG signal, after finding the baseline, it finds a vertical line and extracts the inside of the line as one lead information, so that it is possible to extract the guidance information and segment the region by induction. When the minimum region is extracted, it may be detected how far the graph is from the baseline from the minimum region. The binarization may be performed by displaying the number of pixels up to the graph located above the base line as positive (+) and the number of pixels up to the graph located below the negative as negative (−). The binarization information 910 is an example in which the pixel number is equalized and not displayed, but a method of displaying the pixel number as it is or equalizing the same as in the above example is possible. In addition, the above method corresponds to the case of standard extraction of values, so the minimum interval is available in pixel intervals. By adjusting the horizontal extraction pixel interval according to the purpose of diagnosis and storage, the size of the text file and the desired resolution are free. Can be defined as material.

In addition to the above example, the minimum region 710 may be extracted using a character. This is done by detecting each lead information 420 located above the baseline of the electrocardiograph from the image information 340 before segmentation as shown in FIG. 6. In general, since the lead information is located at a portion corresponding to the start section of each lead, it is generally possible to extract from the baseline portion of the vertical lower portion of the start portion of the lead information to the baseline portion of the vertical lower portion of the start point of the next lead information. It is obvious that the performance is more and more guaranteed when used with the vertical dividing of.

As described above, if the graph area is extracted by the region division technique, all remaining portions may be regarded as text areas. Since the rotation is already corrected in the grid extraction process before the image information is divided into regions, it is obvious that errors in the character recognition process are significantly reduced since it is not necessary to consider the portion of the character rotation in optical character recognition. When various patient information and measurement information are extracted through text recognition, it can be saved as XML-based document or automatically saved in DB, so it can be saved in the database where text and video are stored together with the minimum area extracted image information. When the recording line information is extracted as binary information for each lead as described above, the recording line information can be stored in a text-based database.

1 is a block diagram of a binarization apparatus of an electrocardiogram recording paper according to a preferred embodiment of the present invention.

2 is a flowchart of a binarization processing method of an electrocardiograph recording paper according to a preferred embodiment of the present invention;

3A and 3B illustrate an electrocardiogram recording sheet.

4 is a detailed view of the grid.

5A is a detailed flowchart illustrating preprocessing and grid removing of image information according to an exemplary embodiment of the present invention.

5B-5G illustrate a grid extraction process.

6 is a diagram illustrating a baseline extraction process according to a preferred embodiment of the present invention.

7 is a view showing a minimum region extraction example and a binarized example of recording line information according to a preferred embodiment of the present invention;

Explanation of symbols on the main parts of the drawings

100: control device 102: video reading device (scanner)

104: database 106: display device

108: communication module 110: memory

112: user interface device

Claims (34)

In the binarization method of electrocardiogram recording paper, Preprocessing the electrocardiogram recording paper image information; Removing a grid from the preprocessed image information; Extracting a baseline from the image information from which the grid is removed; Extracting a recording line information (graph) region and a text information (text) region from the image information from which the grid is removed; Recognizing text information from the video information of the text information area; Binarizing the image information of the recording line information area or extracting the minimum area based on the base line; Selectively storing the text information and the binarized recording line information or the text information and the minimum area in a database; Binarization method of the ECG recording paper comprising a. The method of claim 1, The pretreatment is And comparing the horizontal and vertical reference lines of the grid included in the image information with the horizontal and vertical reference lines of the standard grid image information, and rotating the image information corresponding to the standard grid image information. The method according to claim 1 or 2, The pretreatment is And comparing the size of the grid included in the image information with standard grid image information to reduce or enlarge the image information corresponding to the standard grid image information. The method of claim 1, Removal of the grid, And a standard grid image is subtracted from the image information. The method of claim 1, Removal of the grid, And a pixel of a predetermined color is removed from the image information. The method of claim 1, Removal of the grid, And subtracting a standard grid image from the image information and removing pixels of a predetermined color from the image information. The method of claim 1, And extracting the baseline is a line at which a peak of a histogram due to accumulation of brightness information in a horizontal direction (time axis) is located. The method of claim 1, The division of the character information area and the recording line information area is An ECG recording paper binarization method according to claim 1, wherein a character information area and a recording line information area are distinguished based on an internal complexity at each position while moving an image information window having a specific pixel size on the image information. The method of claim 1, The division of the character information area and the recording line information area is ECG recording paper which detects the number of pixels of the recording line farthest from the recording line information from the base line of the image information, sets the vertical range corresponding to the number of pixels as the recording line information area, and sets the remaining part as the character information area. Binarization method. The method of claim 1, The division of the character information area and the recording line information area is After detecting the number of pixels of the recording line farthest from the recording line information from the base line of the image information, setting the vertical range corresponding to the number of pixels as the recording line information area, and determining the remaining part as the character information area, An electrocardiogram, wherein a part of the recording line information area is defined as a character information area based on an internal complexity at each position while moving an image information window having a specific pixel size with respect to the recording line information area; Record paper binarization method. The method according to any one of claims 1 to 10, Electrocardiogram recording paper binarization method characterized in that the recognition of the character information is made through the OCR. The method of claim 1, In the binarization step of the recording line information area, Dividing each lead region based on a division line perpendicular to the base line among the image information of the recording line information region; And binarizing the image information of the region of each lead based on the base line, or extracting a minimum region. The method of claim 1, In the binarization step of the recording line information area, Dividing each lead region based on lead information positioned above the base line among the image information of the recording line information region; And binarizing the image information of the area of each lead based on the base line, or extracting the minimum area. The method of claim 1, In the binarization step of the recording line information area, Dividing each lead region based on a vertical divider line and lead information positioned above the base line among the image information of the recording line information region; And binarizing the image information of the area of each lead based on the base line, or extracting the minimum area. The method of claim 1, wherein the binarization of the recording line information comprises: And detecting the number of pixels from the base line to the position of the recording line information to obtain binary information representing the recording line information. The method of claim 1, The minimum area is, ECG recording paper binarization method characterized in that the image information in the range from the baseline to the recording line located at the maximum distance from the top and bottom, respectively. In the binarization apparatus of an electrocardiogram recording paper, A storage medium storing electrocardiogram recording paper information; Receive and process ECG image data Removing a grid from the preprocessed image information, Extracting a baseline from the image information from which the grid is removed, Extracting a recording line information (graph) region and a text information (text) region from the image information from which the grid is removed, Character information is recognized from the image information of the character information area; Binarizing or extracting a minimum area of the image information of the recording line information area based on the base line; And a control unit for selectively storing the character information and the binarized recording line information or the character information and the minimum area in the storage medium. The method of claim 17, And a scanner or imaging device which is scanned or imaged on the ECG recording paper and provided to the controller, and which is embedded in or remotely connected to the bin of the ECG recording paper. The method of claim 17, And a communication module for transmitting the ECG recording paper information to a terminal connected through a network. And the control unit transmits the electrocardiogram recording sheet information to the corresponding terminal in response to a request for the electrocardiogram recording sheet information through the terminal. The method of claim 17, The pretreatment is And comparing the horizontal and vertical reference lines of the grid included in the image information with the horizontal and vertical reference lines of the standard grid image information, and rotating the image information corresponding to the standard grid image information. The method of claim 17 or 20, The pretreatment is And comparing the size of the grid included in the image information with the size of the standard grid image information, thereby reducing or enlarging the image information corresponding to the standard grid image information. The method of claim 17, Removal of the grid, And the standard grid image is subtracted from the image information. The method of claim 17, Removal of the grid, And the pixel of the predetermined color is removed from the image information. The method of claim 17, Removal of the grid, And a pixel of a predetermined color is removed from the image information while simultaneously removing a standard grid image from the image information. The method of claim 17, The extraction of the base line is an ECG recording paper binarization apparatus, characterized in that the line on which the peak of the histogram due to the accumulation of brightness information in the horizontal direction (time axis) is located. The method of claim 17, The division of the character information area and the recording line information area is And moving the image information window having a specific pixel size on the image information to distinguish between the character information region and the recording line information region based on the internal complexity at each position. The method of claim 17, The division of the character information area and the recording line information area is ECG recording paper which detects the number of pixels of the recording line farthest from the recording line information from the base line of the image information, sets the vertical range corresponding to the number of pixels as the recording line information area, and sets the remaining part as the character information area. Binarization unit. The method of claim 17, The division of the character information area and the recording line information area is After detecting the number of pixels of the recording line farthest from the recording line information from the base line of the image information, setting the vertical range corresponding to the number of pixels as the recording line information area, and determining the remaining part as the character information area, An electrocardiogram, wherein a part of the recording line information area is defined as a character information area based on an internal complexity at each position while moving an image information window having a specific pixel size with respect to the recording line information area; Record paper binarization device. The method according to any one of claims 17 to 28, Electrocardiogram recording paper binarization apparatus, characterized in that the recognition of the character information is made through the OCR. The method of claim 17, The binarization of the recording line information area is Division of each lead region based on a division line perpendicular to the base line among the image information of the recording line information region, Electrocardiogram recording paper binarization apparatus, characterized in that to binarize the image information of the region by the lead or to extract the minimum region on the basis of the base line. The method of claim 17, The binarization of the recording line information area is A region for each lead is divided based on the lead information located above the base line among the image information of the recording line information region, Electrocardiogram recording paper binarization apparatus comprising the step of binarizing the image information of the area of each lead based on the base line, or extracting the minimum area. The method of claim 17, The binarization of the recording line information area is The region for each lead is divided based on the vertical dividing line and the lead information positioned above the base line among the image information of the recording line information region. Electrocardiogram recording paper binarization apparatus, characterized in that to binarize the image information of the region by the lead or to extract the minimum region on the basis of the base line. 18. The method of claim 17, wherein the binarization of the recording line information comprises: And detecting the number of pixels from the base line to the position of the recording line information to obtain binary information representing the recording line information. The method of claim 17, The minimum area is, Electrocardiogram recording paper binarization device characterized in that the image information in the range from the baseline to the recording line located at the maximum distance from the top and bottom, respectively.

Images