WO2021212614A1

WO2021212614A1 - Text error correction method and apparatus, computer-readable storage medium and system

Info

Publication number: WO2021212614A1
Application number: PCT/CN2020/093561
Authority: WO
Inventors: 谢静文; 阮晓雯; 徐亮
Original assignee: 平安科技（深圳）有限公司
Priority date: 2020-04-23
Filing date: 2020-05-30
Publication date: 2021-10-28
Also published as: CN111626118A

Abstract

A text error correction method and apparatus, a system and a computer-readable storage medium, relating to the technology of artificial intelligence. The text error correction method comprises: acquiring an original text image, and preprocessing the original text image to obtain a standard image (S1); performing text recognition on the standard image by using a pre-trained text recognition model to obtain character/word vectors, encoding the character/word vectors to generate key values and corresponding result values, and converting the standard image into an output text according to the key values and the corresponding result values (S2); calculating an editing distance between the output text and a preset standard error correction table by using the key values, and obtaining, according to the editing distance, an error text in the output text and a correct text corresponding to the error text (S3); and replacing the error text with the correct text to obtain a standard output text (S4). The method can solve the problems of low precision and high cost of text error correction. The present invention further relates to a blockchain technology and is also applicable to the field of smart cities.

Description

Text error correction method, device, computer readable storage medium and system

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office, the application number is 202010326324.X, and the invention title is "text error correction method, device, computer-readable storage medium and system" on April 23, 2020. The entire content is incorporated into this application by reference.

Technical field

This application relates to the field of artificial intelligence technology, and in particular to a text error correction method, device, computer-readable storage medium and system.

Background technique

The current text recognition method mostly uses OCR technology to read the text in the image and convert it into a character format that the computer can accept and people can understand. However, because the OCR technology has high requirements on the quality of the input image, a large number of recognition errors are prone to occur in the case of low image accuracy, so it is necessary to perform error correction processing on the recognized characters. The inventor realizes that the traditional method only performs error correction based on the characters in the image information, resulting in that the error correction result directly output by the OCR cannot meet the actual application requirements, and the accuracy rate is low. Therefore, how to achieve low-cost, high-precision text error correction is increasingly being valued.

Summary of the invention

This application provides a text error correction method, device, computer readable storage medium and system, the main purpose of which is to solve the problem of low text error correction accuracy and high cost.

In order to achieve the above objective, a text error correction method provided by this application includes:

Acquiring an original text image, and performing a preprocessing operation on the original text image to obtain a standard image;

Use the pre-trained text recognition model to perform text recognition on the standard image to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values, according to the key values and corresponding Result value, converting the standard image into output text;

Calculate the edit distance between the output text and the preset standard error correction table by using the key value, and obtain the error text in the output text and the correct text corresponding to the error text according to the edit distance;

Replace the error text with the correct text to obtain the standard output text.

In order to solve the above-mentioned problems, the present application also provides a text error correction device, which includes:

The modulation conversion module is used to obtain an original text image, and perform a preprocessing operation on the original text image to obtain a standard image;

The text segmentation module is used to perform text recognition on the standard image using a pre-trained text recognition model to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values, according to all The key value and the corresponding result value, and the standard image is converted into output text;

The distance calculation module is used to calculate the edit distance between the output text and the preset standard error correction table by using the key value, and obtain the error text in the output text and the correctness corresponding to the error text according to the edit distance text;

The error correction output module is used to replace the error text with the correct text to obtain the standard output text.

In order to solve the above-mentioned problems, the present application also provides a computer-readable storage medium with a text error correction program stored on the computer-readable storage medium, and the text error correction program can be executed by one or more processors to achieve The following steps:

In order to solve the above problems, this application also provides a text error correction system, including:

The embodiment of the present application performs a preprocessing operation on the original text image, which removes the disturbing factors in the original image, and provides a pre-foundation for subsequent error correction of the text in the image. Further, compared to the prior art only performing error correction based on the character itself in the image information, the embodiment of the present application calculates the key value of the character and the result value corresponding to the key value, and uses the key value and the result value Compared with a preset standard error correction table, the output text obtained through image recognition technology is corrected to make the correction of errors more accurate. Therefore, the text error correction method, device, and computer-readable storage medium proposed in this application can realize a low-cost, high-precision text error correction solution.

Description of the drawings

FIG. 1 is a schematic flowchart of a text error correction method provided by an embodiment of this application;

2 is a schematic diagram of modules of a text error correction method provided by an embodiment of this application;

3 is a schematic diagram of the internal structure of an electronic device of a text error correction method provided by an embodiment of the application;

The realization, functional characteristics, and advantages of the purpose of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings.

Detailed ways

It should be understood that the specific embodiments described here are only used to explain the present application, and are not used to limit the present application.

This application provides a method for text error correction. Referring to FIG. 1, it is a schematic flowchart of a text error correction method provided by an embodiment of this application. The method can be executed by a device, and the device can be implemented by software and/or hardware.

In this embodiment, the text error correction method includes:

S1. Obtain an original text image, and perform a preprocessing operation on the original text image to obtain a standard image.

In the embodiment of the present application, the original text image is obtained by two-dimensional scanning of paper documents, such as medical invoice paper documents, books, etc.

In order to remove interference factors such as noise in the original text image obtained by two-dimensional scanning, the embodiment of the present application first performs the following preprocessing on the original text image:

Performing enlargement processing on the image signal of the original text image to obtain an enlarged image signal;

Sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.

In detail, the embodiment of the present application utilizes an existing amplifying circuit to amplify the image signal of the original text image. The amplifying circuit is a circuit with a function of amplifying electrical signals composed of a transistor as a control element; other suitable amplifying circuits can be selected according to different amplifying requirements, and the original text image can be amplified without distortion by using the selected amplifying circuit , Get the enlarged image signal.

Further, the embodiment of the present application utilizes an existing sampling circuit to sample the amplified image signal. The sampling circuit is a circuit that can periodically sample the amplified image signal according to a preset sampling frequency.

The embodiment of the present application adopts the above-mentioned enlargement, sampling, and filtering processing on the original text image, removes interference factors such as noise in the original text image, obtains the standard image, and ensures the accuracy of subsequent text error correction.

S2. Use the pre-trained text recognition model to perform text recognition on the standard image to obtain a character/word vector, and encode the character/word vector to generate a key value and a corresponding result value, according to the key value and Corresponding to the result value, the standard image is converted into output text.

Preferably, the text recognition model in the embodiment of the present application may be a pre-trained NER (Named Entity Recognition) model.

Preferably, the NER model adopts the Bi-LSTM-CRF structure, including:

Character/word vector layer: used to convert words and characters in the text contained in the standard image into word/word vectors; Bi-LSTM layer: divide the character/word vector, and divide the character /Word vector encoding to obtain the encoding representation of the character/word vector, and using the encoding representation to label the segmented word/word vector to obtain key values and result values;

CRF layer: splicing key values and result values of the same type, and decoding the spliced text according to the reverse process of encoding to generate the output text.

Wherein, the word/word vector layer uses the trained word vector as an initialization parameter to convert the words and characters in the text contained in the standard image into a word/word vector, and the trained word vector is A set of standard conversion rules summarized in the past when converting word/word vectors.

Since the standard image may contain more text and the sentences in the text may be longer, if only character conversion is performed, the text may be sticky, which is not conducive to subsequent text error correction. Therefore, the embodiment of the application uses The Bi-LSTM layer can segment the character/word vector.

Preferably, the Bi-LSTM layer can use java language to segment the character/word vector, and encode the segmented character/word vector, and the encoding representation includes Key-B, Value-B, Key-I, Value-I, Other-B, Other-I six types of labeling. Among them, Key is the key value, Value is the result value, and Other is the other value.

The CRF layer splices the same type of key value and result value, such as Key-B, Key-I or Value-B, Value-I.

Further, the embodiment of the present application converts the standard image into output text according to the key value and the corresponding result value. For example, in one of the examples of the present application, the standard image contains the text "Pay 2.00 yuan (cash payment) ) The classification is conceited at 0.00 yuan. After processing by the above NER model, the following output text is generated:

Key：{支付,分类自负}Key: {Payment, categorized at your own risk}
Value：{2.00元,0.00元}Value: {2.00 yuan, 0.00 yuan}
Other(现金支付)Other (Cash Payment)

S3. Calculate the edit distance between the output text and the preset standard error correction table by using the key value, and obtain the error text in the output text and the correct text corresponding to the error text according to the edit distance.

Since there may be some content errors in the standard image, such as typos, etc., this embodiment of the present application will use a standard error correction table to correct the above output text.

In the embodiment of the present application, the standard error correction table is composed of a character string without any errors and the key value and result value corresponding to the character string.

The edit distance refers to the minimum number of editing operations required to convert one character string into another character string between two character strings.

For example, calculate the edit distance between the output text ABBD and the character string ABCD in the standard error correction table. Since only the third character in the output text ABBD and the character string ABCD are different, the above method can be used to calculate the minimum number of editing operations as 1, that is, replace the'B' character with the'C' character.

In detail, the embodiment of the present application uses the following edit distance algorithm to calculate the edit distance Sim _topic :

Sim _topic ＝Pearson(R,S)

Wherein, R is the key value of the output text, S is the key value of the standard error correction table, and Pearson is the edit distance calculation.

Further, in order to filter out which character strings in the standard error correction table can be used for error correction of the output text, the embodiment of the present application obtains the error text and the error text in the output text according to the edit distance. The correct text corresponding to the wrong text includes:

Compare the edit distance between the key value of the output text and the key value of the standard error correction table with the preset distance threshold;

When the edit distance is less than the distance threshold, the key value of the corresponding output text is determined to be an error character, and the key value of the corresponding standard error correction table is determined to be the corresponding correct character;

Collecting all the wrong characters to obtain the wrong text in the output text, and collecting the correct characters to obtain the correct text corresponding to the wrong text.

Further, if the edit distance is greater than or equal to the distance threshold, it means that the output text does not match the standard error correction table, and the standard error correction table cannot be used to correct the output text.

S4. Replace the erroneous text with the correct text to obtain standard output text.

In the embodiment of the present application, the correct text can be directly used to replace the erroneous text, so that the error content in the erroneous text can be corrected, and the standard output text can be obtained.

It should be emphasized that, in order to further ensure the privacy and security of the text image, the original text image can also be stored in a node of a blockchain.

The blockchain referred to in this application is a new application mode of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain, essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information for verification. The validity of the information (anti-counterfeiting) and the generation of the next block. The blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.

At the same time, this solution can be tested in the fields of smart medical care in smart cities, so as to promote the construction of smart cities.

The embodiment of the present application performs a preprocessing operation on the original text image, which removes the disturbing factors in the original image, and provides a pre-foundation for subsequent error correction of the text in the image. Further, compared with the prior art that only performs character error correction based on the image information itself, the embodiment of the present application calculates the key value of the character and the result value corresponding to the key value, and uses the key value and the result value and A preset standard error correction table is compared, so that a preset standard error correction table is used to correct the output text obtained through the image recognition technology, so that the error correction is more accurate. Therefore, the text error correction method, device and computer-readable storage medium proposed in this application can realize low-cost, high-precision text error correction.

As shown in Figure 2, it is a functional block diagram of the text error correction device of the present application.

The text error correction device 100 described in this application can be installed in an electronic device. According to the implemented functions, the text error correction device may include an image acquisition module 101, an image segmentation module 102, a matching module 103, and an error correction module 104. The module described in the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

In this embodiment, the functions of each module/unit are as follows:

The modulation conversion module 101 is configured to obtain an original text image, and perform a preprocessing operation on the original text image to obtain a standard image;

The text segmentation module 102 is configured to perform text recognition on the standard image using a pre-trained text recognition model to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values , Converting the standard image into output text according to the key value and the corresponding result value;

The distance calculation module 103 is configured to calculate the edit distance between the output text and a preset standard error correction table by using the key value, and obtain the error text and the error text in the output text according to the edit distance Corresponding correct text;

The error correction output module 104 is configured to replace the error text with the correct text to obtain standard output text.

In detail, the specific implementation steps of each module of the text error correction device 100 are as follows:

The image acquisition module 101 acquires an original text image, and performs a preprocessing operation on the original text image to obtain a standard image.

Sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.

The embodiment of the present application adopts the above-mentioned enlargement, sampling, and filtering processing on the original text image to remove interference factors such as noise in the original text image, obtain the standard image, and ensure the accuracy of subsequent text error correction.

The image segmentation module 102 uses a pre-trained text recognition model to perform text recognition on the standard image to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values, according to all The key value and the corresponding result value are converted, and the standard image is converted into output text.

Since the standard image may contain a lot of text and the sentences in the text may be longer, if the character conversion is performed directly, the text may be sticky, which is not conducive to subsequent text error correction. Therefore, the embodiment of the application uses The text recognition model performs text recognition and segmentation processing on the standard image.

Preferably, the NER model adopts the Bi-LSTM-CRF structure, including:

Character/word vector layer: used to convert words and characters in the text contained in the standard image to obtain a word/word vector;

Bi-LSTM layer: used to segment the character/word vector, encode the segmented character/word vector to obtain the encoding representation of the character/word vector, and use the encoding representation to /Word vector for labeling, get key value and result value;

The Bi-LSTM layer may use java language to encode the word/word vector, and the encoding representation includes six types: Key-B, Value-B, Key-I, Value-I, Other-B, Other-I Label type. Among them, Key is the key value, Value is the result value, and Other is the other value.

Further, the embodiment of the present application converts the standard image into output text according to the key value and the corresponding result value. For example, in one of the examples of the present application, the standard image has the text "Pay 2.00 yuan (cash payment) ) The classification is conceited at 0.00 yuan. After processing by the above NER model, the following output text is generated:

Key：{支付,分类自负}Key: {Payment, categorized at your own risk}
Value：{2.00元,0.00元}Value: {2.00 yuan, 0.00 yuan}

Other (Cash Payment)

The matching module 103 uses the key value to calculate the edit distance between the output text and the preset standard error correction table, and obtains the error text in the output text and the correct text corresponding to the error text according to the edit distance .

In the embodiment of the present application, the standard error correction table is composed of a character string without any errors and the key value and result value corresponding to the character string. The edit distance refers to the minimum number of editing operations required to convert one character string into another character string between two character strings.

Sim _topic ＝Pearson(R,S)

The error correction module 104 replaces the error text with the correct text to obtain the standard output text.

As shown in FIG. 3, it is a schematic diagram of the structure of an electronic device implementing the text error correction method of the present application.

The electronic device 1 may include a processor 10, a memory 11, and a bus, and may also include a computer program stored in the memory 11 and running on the processor 10, such as a text error correction program 12.

Wherein, the memory 11 includes at least one type of readable storage medium, the readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (for example: SD or DX memory, etc.), magnetic memory, magnetic disk, CD etc. The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, for example, a mobile hard disk of the electronic device 1. In other embodiments, the memory 11 may also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a smart media card (SMC), and a secure digital (Secure Digital) equipped on the electronic device 1. , SD) card, flash card (Flash Card), etc. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 can be used not only to store application software and various data installed in the electronic device 1, such as the code of the text error correction program 12, etc., but also to temporarily store data that has been output or will be output. The computer-readable storage medium may be non-volatile or volatile.

The processor 10 may be composed of integrated circuits in some embodiments, for example, may be composed of a single packaged integrated circuit, or may be composed of multiple integrated circuits with the same function or different functions, including one or more Combinations of central processing unit (CPU), microprocessor, digital processing chip, graphics processor, and various control chips, etc. The processor 10 is the control unit of the electronic device, which uses various interfaces and lines to connect the various components of the entire electronic device, and runs or executes programs or modules stored in the memory 11 (for example, executing Text error correction programs, etc.), and call data stored in the memory 11 to execute various functions of the electronic device 1 and process data.

The bus may be a peripheral component interconnect standard (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus and so on. The bus is configured to implement connection and communication between the memory 11 and at least one processor 10 and the like.

FIG. 3 only shows an electronic device with components. Those skilled in the art can understand that the structure shown in FIG. 3 does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure. Components, or a combination of certain components, or different component arrangements.

For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to various components. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, thereby controlling power The device implements functions such as charge management, discharge management, and power consumption management. The power supply may also include any components such as one or more DC or AC power supplies, recharging devices, power failure detection circuits, power converters or inverters, and power status indicators. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

Further, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is usually used in the electronic device 1 Establish a communication connection with other electronic devices.

Optionally, the electronic device 1 may also include a user interface. The user interface may be a display (Display) and an input unit (such as a keyboard (Keyboard)). Optionally, the user interface may also be a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display can also be appropriately called a display screen or a display unit, which is used to display the information processed in the electronic device 1 and to display a visualized user interface.

It should be understood that the embodiments are only for illustrative purposes, and are not limited by this structure in the scope of the patent application.

The text error correction program 12 stored in the memory 11 in the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can realize:

Use the pre-trained text recognition model to perform text recognition and segmentation processing on the standard image, and generate key values and corresponding result values for the standard image after segmentation. According to the key values and corresponding result values, The standard image is converted into output text;

Further, if the integrated module/unit of the electronic device 1 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory) .

In the several embodiments provided in this application, it should be understood that the disclosed equipment, device, and method may be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the objectives of the solutions of the embodiments.

In addition, the functional modules in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware, or may be implemented in the form of hardware plus software functional modules.

For those skilled in the art, it is obvious that the present application is not limited to the details of the foregoing exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or basic characteristics of the application.

Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of this application is defined by the appended claims rather than the above description, and therefore it is intended to fall into the claims. All changes in the meaning and scope of the equivalent elements of are included in this application. Any associated diagram marks in the claims should not be regarded as limiting the claims involved.

In addition, it is obvious that the word "including" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claims can also be implemented by one unit or device through software or hardware. The second class words are used to indicate names, and do not indicate any specific order.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the application and not to limit them. Although the application has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the application can be Make modifications or equivalent replacements without departing from the spirit and scope of the technical solution of the present application.

Claims

A method for text error correction, wherein the method includes:

Acquiring an original text image, and performing a preprocessing operation on the original text image to obtain a standard image;

Use the pre-trained text recognition model to perform text recognition on the standard image to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values, according to the key values and corresponding Result value, converting the standard image into output text;

Calculate the edit distance between the output text and the preset standard error correction table by using the key value, and obtain the error text in the output text and the correct text corresponding to the error text according to the edit distance;

Replace the error text with the correct text to obtain the standard output text.
5. The text error correction method according to claim 1, wherein the preprocessing operation of the original text image to obtain a standard image comprises:

Performing enlargement processing on the image signal of the original text image to obtain an enlarged image signal;

Sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.
The text error correction method according to claim 1, wherein the text recognition model comprises:

The word/word vector layer is used to convert words and characters in the text contained in the standard image to obtain a word/word vector;

The Bi-LSTM layer is used to divide the character/word vector, encode the character/word vector after the segmentation, to obtain the coding representation of the character/word vector, and use the coding representation for the character /Word vector for labeling, get key value and result value;

The CRF layer is used to splice the key values and result values of the same type, and decode the spliced text according to the reverse process of encoding to generate the output text.
5. The text error correction method according to claim 3, wherein said calculating the edit distance between the output text and a preset standard error correction table comprises:

The edit distance is calculated using the following edit distance algorithm:

Sim topic ＝Pearson(R,S)

Where R is the key value of the output text, S is the key value of the standard error correction table, Pearson is the edit distance calculation, and Sim topic is the edit distance between the key values.
5. The text error correction method according to claim 4, wherein said obtaining the error text in the output text and the correct text corresponding to the error text according to the edit distance comprises:

Compare the edit distance between the key value of the output text and the key value of the standard error correction table with the preset distance threshold;

When the edit distance is less than the distance threshold, the key value of the corresponding output text is determined to be an error character, and the key value of the corresponding standard error correction table is determined to be the corresponding correct character;

Collecting all the wrong characters to obtain the wrong text in the output text, and collecting the correct characters to obtain the correct text corresponding to the wrong text.
A text error correction device, wherein the device includes:

The modulation conversion module is used to obtain an original text image, and perform a preprocessing operation on the original text image to obtain a standard image;

The text segmentation module is used to perform text recognition on the standard image using a pre-trained text recognition model to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values, according to all The key value and the corresponding result value, and the standard image is converted into output text;

The distance calculation module is used to calculate the edit distance between the output text and the preset standard error correction table by using the key value, and obtain the error text in the output text and the correctness corresponding to the error text according to the edit distance text;

The error correction output module is used to replace the error text with the correct text to obtain the standard output text.
7. The text error correction device according to claim 6, wherein the preprocessing operation of the original text image to obtain a standard image comprises:

Performing enlargement processing on the image signal of the original text image to obtain an enlarged image signal;

Sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.
7. The text error correction device of claim 6, wherein the text recognition model comprises:

The word/word vector layer is used to convert words and characters in the text contained in the standard image to obtain a word/word vector;

The Bi-LSTM layer is used to divide the character/word vector, encode the character/word vector after the segmentation, to obtain the coding representation of the character/word vector, and use the coding representation for the character /Word vector for labeling, get key value and result value;

The CRF layer is used to splice the key values and result values of the same type, and decode the spliced text according to the reverse process of encoding to generate the output text.
8. The text error correction device according to claim 8, wherein said calculating the edit distance between the output text and a preset standard error correction table comprises:

The edit distance is calculated using the following edit distance algorithm:

Sim topic ＝Pearson(R,S)

Where R is the key value of the output text, S is the key value of the standard error correction table, Pearson is the edit distance calculation, and Sim topic is the edit distance between the key values.
7. The text error correction device according to claim 6, wherein when the modulation conversion module performs a preprocessing operation on the original text image, it executes:

Performing enlargement processing on the image signal of the original text image to obtain an enlarged image signal;

Sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.
A computer-readable storage medium, wherein a text error correction program is stored on the computer-readable storage medium, and the text error correction program can be executed by one or more processors to implement the following steps:

Acquiring an original text image, and performing a preprocessing operation on the original text image to obtain a standard image;

Use the pre-trained text recognition model to perform text recognition on the standard image to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values, according to the key values and corresponding Result value, converting the standard image into output text;

Calculate the edit distance between the output text and the preset standard error correction table by using the key value, and obtain the error text in the output text and the correct text corresponding to the error text according to the edit distance;

Replace the error text with the correct text to obtain the standard output text.
11. The computer-readable storage medium of claim 11, wherein the preprocessing operation of the original text image to obtain a standard image comprises:

Performing enlargement processing on the image signal of the original text image to obtain an enlarged image signal;

Sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.
11. The computer-readable storage medium of claim 11, wherein the text recognition model comprises:

The word/word vector layer is used to convert words and characters in the text contained in the standard image to obtain a word/word vector;

The Bi-LSTM layer is used to divide the character/word vector, encode the character/word vector after the segmentation, to obtain the coding representation of the character/word vector, and use the coding representation for the character /Word vector for labeling, get key value and result value;

The CRF layer is used to splice the key values and result values of the same type, and decode the spliced text according to the reverse process of encoding to generate the output text.
15. The computer-readable storage medium of claim 13, wherein the calculating the edit distance between the output text and a preset standard error correction table comprises:

The edit distance is calculated using the following edit distance algorithm:

Sim topic ＝Pearson(R,S)

Where R is the key value of the output text, S is the key value of the standard error correction table, Pearson is the edit distance calculation, and Sim topic is the edit distance between the key values.
14. The computer-readable storage medium of claim 14, wherein the obtaining the error text in the output text and the correct text corresponding to the error text according to the edit distance comprises:

Compare the edit distance between the key value of the output text and the key value of the standard error correction table with the preset distance threshold;

When the edit distance is less than the distance threshold, the key value of the corresponding output text is determined to be an error character, and the key value of the corresponding standard error correction table is determined to be the corresponding correct character;

Collecting all the wrong characters to obtain the wrong text in the output text, and collecting the correct characters to obtain the correct text corresponding to the wrong text.
A text error correction system, wherein the text error correction system includes:

The modulation conversion module is used to obtain an original text image, and perform a preprocessing operation on the original text image to obtain a standard image;

The text segmentation module is used to perform text recognition on the standard image using a pre-trained text recognition model to obtain a character/word vector, and encode the character/word vector to generate key values and corresponding result values, according to all The key value and the corresponding result value, and the standard image is converted into output text;

The distance calculation module is used to calculate the edit distance between the output text and the preset standard error correction table by using the key value, and obtain the error text in the output text and the correctness corresponding to the error text according to the edit distance text;

The error correction output module is used to replace the error text with the correct text to obtain the standard output text.
The text error correction system according to claim 16, wherein the preprocessing operation of the original text image to obtain a standard image comprises:

Performing enlargement processing on the image signal of the original text image to obtain an enlarged image signal;

Sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.
The text error correction system according to claim 16, wherein the text recognition model comprises:

The word/word vector layer is used to convert words and characters in the text contained in the standard image to obtain a word/word vector;

The Bi-LSTM layer is used to divide the character/word vector, encode the character/word vector after the segmentation, to obtain the coding representation of the character/word vector, and use the coding representation for the character /Word vector for labeling, get key value and result value;

The CRF layer is used to splice the key values and result values of the same type, and decode the spliced text according to the reverse process of encoding to generate the output text.
The text error correction system according to claim 18, wherein said calculating the edit distance between the output text and a preset standard error correction table comprises:

The edit distance is calculated using the following edit distance algorithm:

Sim topic ＝Pearson(R,S)

Where R is the key value of the output text, S is the key value of the standard error correction table, Pearson is the edit distance calculation, and Sim topic is the edit distance between the key values.
The text error correction system according to claim 16, wherein when the modulation conversion module performs a preprocessing operation on the original text image, it executes:

Performing amplification processing on the image signal of the original text image to obtain an amplified image signal; sampling the amplified image signal to obtain a sampling signal;

Filtering the sampled signal to obtain the standard image.