CN111832371A - Text and picture correction method, device, electronic device and machine-readable storage medium - Google Patents

Abstract

Embodiments of the present invention provide a text picture correction method, device, electronic device, and machine-readable storage medium. The method includes: acquiring a text picture to be corrected; performing text line detection on the text picture to obtain each text line of the text picture The specified point position information of multiple characters in each text line; according to the specified point position information of multiple characters in each text line, the distortion curve of each text line is calculated; based on the distortion curve of each text line, the text picture is corrected and mapped, and the corrected text picture. By applying the embodiments of the present invention, the distorted text content in the text picture can be corrected, and the user's reading experience can be improved.

Description

Text and picture correction method, device, electronic device and machine-readable storage medium

technical field

The present invention relates to the technical field of image processing, and in particular, to a text and picture correction method, device, electronic device and machine-readable storage medium.

Background technique

With the development of electronic technology, it is often necessary to photograph or scan paper texts to meet people's needs for easy viewing of texts and sharing of text content. However, when photographing or scanning paper texts, due to the thickness of books and other carriers, it is unavoidable that the text content in the photographed or scanned text images will be distorted, affecting people's reading experience.

SUMMARY OF THE INVENTION

The purpose of the embodiments of the present invention is to provide a text picture correction method, device, electronic device and machine-readable storage medium, so as to correct the distorted text content in the text picture and improve the reading experience. The specific technical solutions are as follows:

In a first aspect, an embodiment of the present invention provides a text picture correction method, the method comprising:

Get the text image to be corrected;

Perform text line detection on the text picture, and obtain designated point position information of multiple characters in each text line of the text picture;

Calculate the distortion curve of each text line according to the specified point position information of the multiple characters in each text line;

Correction mapping is performed on the text picture based on the distortion curve of each text line to obtain a corrected text picture.

Preferably, the text line detection is performed on the text picture to obtain the specified point position information of multiple characters in each text line of the text picture, including:

Perform document area detection on the text picture to obtain the coordinates of the detection frame of the document area in the text picture;

According to the coordinates of the detection frame, the specified point position information of a plurality of characters in each text line of the document area is detected and obtained from the document area.

Preferably, calculating the distortion curve of each text line according to the specified point position information of multiple characters in each text line includes:

Build an initialization curve;

Calculate the error value between the first ordinate in the specified point position information and the second ordinate on the initialization curve according to the specified point position information of the characters in each text line, and construct each text line Corresponding to the error function of the initialization curve, the second ordinate is the ordinate corresponding to the same abscissa in the specified position information on the initialization curve;

According to the error function of each text line corresponding to the initialization curve, through curve fitting, the distortion curve of each text line is obtained.

Preferably, according to the error function of each text line corresponding to the initialization curve, through curve fitting, the distortion curve of each text line is obtained, including:

A minimization operation is performed on the error function of each text line corresponding to the initialization curve, and the distortion curve of each text line is obtained by fitting.

Preferably, the correction mapping is performed on the text picture based on the distortion curve of each text line to obtain a corrected text picture, including:

Composing the twisted curves of each text line into a three-dimensional grid to form a twisted surface;

The distorted surface is mapped to a flat picture to obtain a corrected text picture.

Preferably, after the text picture is corrected and mapped based on the distortion curve of each text line to obtain the corrected text picture, the method further includes:

Character recognition technology is used to recognize each character in the corrected text picture.

In a second aspect, an embodiment of the present invention provides a text picture correction device, including:

The acquisition module is used to acquire the text image to be corrected;

a detection module, configured to perform text line detection on the text picture, and obtain designated point position information of multiple characters in each text line of the text picture;

a calculation module, configured to calculate the distortion curve of each text line according to the specified point position information of a plurality of characters in each text line;

The correction module is configured to perform correction mapping on the text picture based on the distortion curve of each text line to obtain a corrected text picture.

Preferably, the detection module is specifically used for:

Perform document area detection on the text picture to obtain the coordinates of the detection frame of the document area in the text picture;

According to the coordinates of the detection frame, the specified point position information of a plurality of characters in each text line of the document area is detected and obtained from the document area.

Preferably, the computing module is specifically used for:

Build an initialization curve;

Calculate the error value between the first ordinate in the specified point position information and the second ordinate on the initialization curve according to the specified point position information of the characters in each text line, and construct each text line Corresponding to the error function of the initialization curve, the second ordinate is the ordinate corresponding to the same abscissa in the specified position information on the initialization curve;

According to the error function of each text line corresponding to the initialization curve, through curve fitting, the distortion curve of each text line is obtained.

Preferably, when the calculation module is used to obtain the distortion curve of each text line through curve fitting according to the error function corresponding to the initialization curve of each text line, it is specifically used for:

A minimization operation is performed on the error function of each text line corresponding to the initialization curve, and the distortion curve of each text line is obtained by fitting.

Preferably, the correction module is specifically used for:

Composing the twisted curves of each text line into a three-dimensional grid to form a twisted surface;

The distorted surface is mapped to a flat picture to obtain a corrected text picture.

Preferably, the device further comprises:

The character recognition module is used for using character recognition technology to recognize each character in the corrected text picture.

In a third aspect, an embodiment of the present invention provides an electronic device, including a processor and a memory, wherein,

the memory for storing computer programs;

The processor is configured to implement the method provided by the first aspect of the embodiment of the present invention when executing the computer program stored in the memory.

In a fourth aspect, an embodiment of the present invention provides a machine-readable storage medium, where a computer program is stored in the machine-readable storage medium, and when the computer program is executed by a processor, the first aspect of the embodiment of the present invention provides Methods.

A text picture correction method, device, electronic device, and machine-readable storage medium provided by the embodiments of the present invention acquire a text picture to be corrected, perform text line detection on the text picture, and obtain a plurality of characters in each text line of the text picture According to the specified point position information of each text line, the distortion curve of each text line is calculated according to the specified point position information of multiple characters in each text line, and based on the distortion curve of each text line, the text image is corrected and mapped to obtain the corrected text image. By detecting the text lines in the text picture, the designated point position information of multiple characters in each text line is obtained. The designated point position information of the character represents the specific position of the character in the text image. The position information can calculate the distortion curve of each text line. The distortion curve represents the distortion of the text line. Based on the distortion of each text line, the original distorted text image can be corrected into an undistorted text image through correction mapping. Correction of the distorted text content in the text picture is realized, and the user's reading experience is improved.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

1 is a schematic flowchart of a text and picture correction method according to an embodiment of the present invention;

2 is a schematic diagram of a twisted curved surface according to an embodiment of the present invention;

Fig. 3 (a) is the display schematic diagram of the picture to be corrected according to the embodiment of the present invention;

Fig. 3 (b) is the display schematic diagram of the corrected picture according to the embodiment of the present invention;

4 is a schematic flowchart of text picture character recognition according to an embodiment of the present invention;

Fig. 5 (a) is the display schematic diagram of the picture to be corrected according to the embodiment of the present invention;

FIG. 5(b) is a schematic display diagram of a corrected picture according to an embodiment of the present invention;

6 is a schematic structural diagram of a text and picture correction device according to an embodiment of the present invention;

7 is a schematic structural diagram of an electronic device according to an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a display system according to an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

In order to correct the distorted text content in the text picture and improve the user's reading experience, the embodiments of the present invention provide a text picture correction method, device, electronic device, and machine-readable storage medium. Below, the text picture correction method provided by the embodiment of the present invention is first introduced.

The execution subject of the text picture correction method provided by the embodiment of the present invention may be a terminal device with a shooting function, such as a mobile phone and a camera; it may also be a server in the background, and after the text picture is captured by the image acquisition device, it is sent to the background server for correction; It can also be a display device, after the text picture captured by the image acquisition device is sent to the display device for correction and display. The manner of implementing the text and picture correction method provided by the embodiment of the present invention may be at least one manner of software, hardware circuit and logic circuit provided in the execution body.

As shown in FIG. 1 , a schematic flowchart of a text and picture correction method provided by an embodiment of the present invention, the method may include the following steps:

S101, acquiring a text picture to be corrected.

The text image to be corrected is a text image in which the text content is distorted in the included image content. The execution body of the embodiment of the present invention may, when receiving the correction request input by the user, obtain the text picture to be corrected specified by the user from the storage medium to perform the correction operation; When there is distortion in the text picture, start to acquire the text picture to be corrected and perform the correction operation.

S102: Perform text line detection on the text picture to obtain designated point position information of multiple characters in each text line of the text picture.

For text pictures, the traditional text line detection technology can be used to detect each text line in the text picture. The traditional text line detection technology mainly detects characters. After detecting the characters in each text line, input The position information of the specified point of the character, the position information of the specified point of the character can be determined by the setting of the algorithm, which can be the position information of the bottom middle point of the character, the position information of the top middle point, the position information of the center point of the character, and so on.

Due to the influence of the shooting angle, focus range, etc., when taking a text picture, the text picture may not only include the area where the text document is located, but may also include content in other non-document areas. In order to improve the efficiency of text line detection and reduce the influence of the content of the non-document area on the text line detection result, preferably, S102 can be specifically implemented by the following steps:

The first step is to perform document area detection on the text image to obtain the coordinates of the detection frame of the document area in the text image.

In the second step, according to the coordinates of the detection frame, from the document area, the designated point position information of the multiple characters in each text line in the document area is detected and obtained.

For the acquired text image to be corrected, firstly, the document area detection can be performed on the text image, the document area in the text image can be detected, and the document area detection method can be realized by using a feature matching algorithm, a deep learning algorithm, etc., in order to ensure The detection accuracy and computational efficiency can usually be performed by selecting a deep learning algorithm for document area detection. Through the detection, the detection frame coordinates of the document area can be obtained. Neural Network, CNN), Recurrent Neural Network (Recurrent Neural Network, RNN) and other algorithms, the algorithm used can identify the document content in the text image, and there is no detailed limitation here.

The coordinates of the detection frame can be the coordinates of the four vertices of the detection frame where the document area is located, which represent the specific position of the detection frame in the text picture. The distance of the box is converted to the specified point position information of the character. Exemplarily, with the lower left corner of the detection frame as the origin, the specified point position information is marked according to the distance between the position of the specified point of the character relative to the position of the lower left corner of the detection frame.

In the specific text line detection method, the connected domain recognition method can be used to identify possible character candidate boxes, and at the same time, the average height of all characters is calculated according to the recognition results of the character candidate boxes, and some characters that are too large and too small are filtered out. Candidate frame, output the position information of the center point of the bottom edge of the candidate frame of each remaining character as the specified point position information of the character.

S103: Calculate the distortion curve of each text line according to the designated point position information of the multiple characters in each text line.

After detecting and obtaining the designated point position information of multiple characters in each text line, the designated point position information of the character represents the position of the character in the text line, by connecting the designated points of multiple characters in a text line, or , and further, by fitting the connection lines of the specified points of multiple characters in a text line, the distortion curve of the text line can be obtained. The distortion curve represents the distortion of the text line. The larger the distortion curve fluctuation, the The more distorted the lines of text are.

If the designated point position information of multiple characters in each text line is detected and obtained, the designated points of multiple characters in one text line are directly connected as a twisted curve, limited by the connection rules, the twisted curve and the actual text line There is often an error between distorted conditions. In order to obtain the distortion curve of each text line more accurately, preferably, S103 can be specifically implemented by the following steps:

The first step is to construct an initialization curve.

The second step is to calculate the error value between the first ordinate in the specified point position information and the second ordinate on the initialization curve according to the position information of the designated points of the characters in each text line, and construct each text line corresponding to the initialization curve The error function of , wherein the second ordinate is the ordinate corresponding to the same abscissa in the specified position information on the initialization curve.

In the third step, according to the error function of each text line corresponding to the initialization curve, through curve fitting, the distortion curve of each text line is obtained.

First, a curve can be initialized as a reference line for fitting the twisted curve. In a special case, the initialization curve can also be a straight line. The initialization curve is used as a reference line, and the error between the position information of the specified points of multiple characters in the text line and the coordinates of the corresponding points on the initialization curve can be obtained. The corresponding relationship can be the same abscissa, and the error is the same abscissa. The error in the y direction of the character and the initialization curve. For a text line, the errors corresponding to multiple characters on the text line can be obtained, and an error function can be further constructed. Based on the error function, the distortion curve of each text line can be obtained by curve fitting.

Optionally, according to the error function of each text line corresponding to the initialization curve, the step of obtaining the distortion curve of each text line through curve fitting may specifically be: performing a minimization operation on the error function of each text line corresponding to the initialization curve , and the distortion curve of each text line is obtained by fitting.

Specifically, the distortion curve of the text line can be represented by a one-variable multi-degree equation (such as a one-variable quadratic equation), and for the error function corresponding to each text line, by minimizing the error function, the one-variable multi-dimensional equation can be calculated. The parameters of the sub-equation can be accurately determined, and the distortion curve can be obtained by fitting, which is used to characterize the distortion of each text line.

S104 , correcting and mapping the text picture based on the distortion curve of each text line to obtain a corrected text picture.

After calculating the distortion curve of each text line, if the distortion curve of each text line is straightened and tiled, the corrected effect of each text line can be mapped. However, due to the distortion of the entire text image It needs to be limited by the distortion of all text lines. If each text line is corrected separately, the effect after correction will be less than ideal, and it is easy to display irregular characters. Therefore, under normal circumstances, a 3D mapping method can be selected to perform the correction mapping of the text picture, and specifically, the mapping can be realized by a 3D mapping function.

Preferably, S104 can be specifically implemented by the following steps:

The distorted curves of each text line are formed into a three-dimensional grid to form a distorted surface; the distorted surface is mapped to a plane image to obtain a corrected text image.

Based on the twisted curve of each text line, a twisted surface as shown in Figure 2 is formed through a three-dimensional mesh. The process of correcting and mapping the distorted surface can be implemented by remapping, such as the CV2.remap image correction method. Remapping is the process of mapping a pixel at a location in one image to a specified location in another image. In this way, the deformed texture corresponding to the three-dimensional distorted surface we composed can be mapped to the plane image, so as to obtain the corrected text image.

Fig. 3(a) shows the obtained original text picture. By using the method provided by the embodiment of the present invention, the text picture shown in Fig. 3(a) is corrected, and the result shown in Fig. 3(b) can be achieved. Correction effect, it can be seen that Figure 3(b) achieves the ideal correction effect, and the text content is tiled and corrected, making the user more comfortable and intuitive to read.

By applying the embodiment of the present invention, a text image to be corrected is obtained, and text line detection is performed on the text image to obtain designated point position information of multiple characters in each text line of the text image, and according to the designated point positions of multiple characters in each text line information, the distortion curve of each text line is calculated, and based on the distortion curve of each text line, the text image is corrected and mapped to obtain the corrected text image. By detecting the text lines in the text picture, the designated point position information of multiple characters in each text line is obtained. The designated point position information of the character represents the specific position of the character in the text image. The position information can calculate the distortion curve of each text line. The distortion curve represents the distortion of the text line. Based on the distortion of each text line, the original distorted text image can be corrected into an undistorted text image through correction mapping. Correction of the distorted text content in the text picture is realized, and the user's reading experience is improved.

Based on the embodiment shown in FIG. 1 , after correcting the text picture in this embodiment of the present invention, character recognition can also be performed on the text picture, as shown in FIG. 4 , the following steps may be included:

S401, acquiring a text picture to be corrected.

S402: Perform text line detection on the text picture to obtain designated point position information of multiple characters in each text line of the text picture.

S403: Calculate the distortion curve of each text line according to the designated point position information of the multiple characters in each text line.

S404 , correcting and mapping the text picture based on the distortion curve of each text line to obtain a corrected text picture.

The steps from S401 to S404 in this embodiment of the present invention are the same as the steps from S101 to S104 in the embodiment shown in FIG. 1 , and details are not repeated here.

S405, using character recognition technology to recognize each character in the corrected text picture.

For traditional character recognition technology, such as Optical Character Recognition (OCR) technology, because the text image to be corrected has text bending, it is difficult to accurately recognize the character during character recognition, or the recognition result is also It may be wrong. As shown in Figure 5(a), when performing character recognition, the actual difference between the recognition frame and the character is large, resulting in the character not being recognized, or the recognition being wrong.

In order to improve the accuracy of the character recognition result, after performing the steps of correcting the text picture provided by the embodiment of the present invention, a character recognition technology such as OCR can be used to recognize the characters in the corrected text picture. The text picture is a picture without bending or a picture with little bending, and then using character recognition technology for character recognition, the characters in the picture can be accurately recognized, as shown in Figure 5(b), during character recognition, the recognition box Each corresponding text line can be covered more accurately, so that the characters in the picture can be accurately recognized.

Applying the above method embodiment, the embodiment of the present invention further provides a text and picture correction device. As shown in FIG. 6 , the text and picture correction device may include:

an acquisition module 610, configured to acquire the text picture to be corrected;

A detection module 620, configured to perform text line detection on the text picture, and obtain designated point position information of multiple characters in each text line of the text picture;

a calculation module 630, configured to calculate the distortion curve of each text line according to the specified point position information of multiple characters in each text line;

The correction module 640 is configured to perform correction mapping on the text picture based on the distortion curve of each text line to obtain a corrected text picture.

Preferably, the detection module 620 can be specifically used for:

Perform document area detection on the text picture to obtain the coordinates of the detection frame of the document area in the text picture;

According to the coordinates of the detection frame, the specified point position information of a plurality of characters in each text line of the document area is detected and obtained from the document area.

Preferably, the computing module 630 can be specifically used for:

Build an initialization curve;

Calculate the error value between the first ordinate in the specified point position information and the second ordinate on the initialization curve according to the specified point position information of the characters in each text line, and construct each text line Corresponding to the error function of the initialization curve, the second ordinate is the ordinate corresponding to the same abscissa in the specified position information on the initialization curve;

According to the error function of each text line corresponding to the initialization curve, through curve fitting, the distortion curve of each text line is obtained.

Preferably, when the calculation module 630 is used to obtain the distortion curve of each text line through curve fitting according to the error function corresponding to the initialization curve of each text line, it can also be used for :

A minimization operation is performed on the error function of each text line corresponding to the initialization curve, and the distortion curve of each text line is obtained by fitting.

Preferably, the correction module 640 can be specifically used for:

Composing the twisted curves of each text line into a three-dimensional grid to form a twisted surface;

The distorted surface is mapped to a flat picture to obtain a corrected text picture.

Preferably, the device may also include:

The character recognition module is used for using character recognition technology to recognize each character in the corrected text picture.

By applying the embodiment of the present invention, a text image to be corrected is obtained, and text line detection is performed on the text image to obtain designated point position information of multiple characters in each text line of the text image, and according to the designated point positions of multiple characters in each text line information, the distortion curve of each text line is calculated, and based on the distortion curve of each text line, the text image is corrected and mapped to obtain the corrected text image. By detecting the text lines in the text picture, the designated point position information of multiple characters in each text line is obtained. The designated point position information of the character represents the specific position of the character in the text image. The position information can calculate the distortion curve of each text line. The distortion curve represents the distortion of the text line. Based on the distortion of each text line, the original distorted text image can be corrected into an undistorted text image through correction mapping. Correction of the distorted text content in the text picture is realized, and the user's reading experience is improved.

An embodiment of the present invention further provides an electronic device, as shown in FIG. 7 , including a processor 701 and a memory 702, wherein,

The memory 702 is used to store computer programs;

The processor 701 is configured to implement any of the above text and picture correction methods when executing the computer program stored in the memory.

Data transmission can be performed between the memory 702 and the processor 701 through a wired connection or a wireless connection, and the electronic device can communicate with other devices through a wired communication interface or a wireless communication interface.

The above-mentioned memory may include random access memory (Random Access Memory, RAM), and may also include non-volatile memory (Non-Volatile Memory, NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

The above-mentioned processor may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; it may also be a digital signal processor (Digital Signal Processing, DSP), an application-specific integrated circuit ( Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

In this embodiment, by reading the machine-executable instructions stored in the machine-readable storage medium, and by running the machine-executable instructions, the processor of the electronic device can achieve: acquiring the text image to be corrected, and performing processing on the text image. Text line detection, obtain the specified point position information of multiple characters in each text line of the text image, calculate the twist curve of each text line according to the specified point position information of multiple characters in each text line, and based on the twist curve of each text line , perform correction mapping on the text image to obtain the corrected text image. By detecting the text lines in the text picture, the designated point position information of multiple characters in each text line is obtained. The designated point position information of the character represents the specific position of the character in the text image. The position information can calculate the distortion curve of each text line. The distortion curve represents the distortion of the text line. Based on the distortion of each text line, the original distorted text image can be corrected into an undistorted text image through correction mapping. Correction of the distorted text content in the text picture is realized, and the user's reading experience is improved.

In addition, the present invention also provides a computer-readable storage medium, where a computing program is stored in the computer-readable storage medium, and when the computing program is executed by a processor, any one of the above text and picture correction methods is implemented.

In this embodiment, the machine-readable storage medium stores a machine computer program that executes the text picture correction method provided by the embodiment of the present invention at runtime, so it can realize: acquiring the text picture to be corrected, and performing text line detection on the text picture , obtain the designated point position information of multiple characters in each text line of the text image, calculate the distortion curve of each text row according to the designated point position information of multiple characters in each text row, and based on the distortion curve of each text row, analyze the text The picture is corrected and mapped to obtain a corrected text picture. By detecting the text lines in the text picture, the designated point position information of multiple characters in each text line is obtained. The designated point position information of the character represents the specific position of the character in the text image. The position information can calculate the distortion curve of each text line. The distortion curve represents the distortion of the text line. Based on the distortion of each text line, the original distorted text image can be corrected into an undistorted text image through correction mapping. Correction of the distorted text content in the text picture is realized, and the user's reading experience is improved.

An embodiment of the present invention further provides a display system, as shown in FIG. 8 , including a display screen 801 and an electronic device 802, wherein,

The display screen 801 is used to display the text picture to be corrected;

The electronic device 802 is used to implement the following steps:

Get the text image to be corrected;

Perform text line detection on the text picture, and obtain designated point position information of multiple characters in each text line of the text picture;

Calculate the distortion curve of each text line according to the specified point position information of the multiple characters in each text line;

Correction mapping is performed on the text picture based on the distortion curve of each text line to obtain a corrected text picture.

Optionally, the above electronic device 802 may also implement any of the above text and picture correction methods when running.

By applying the embodiment of the present invention, by acquiring the text picture to be corrected, the text line detection is performed on the text picture, so as to obtain the designated point position information of multiple characters in each text line of the text image, according to the designated point position information of the multiple characters in each text line. The position information is used to calculate the distortion curve of each text line, and based on the distortion curve of each text line, the text image is corrected and mapped to obtain the corrected text image. By detecting the text lines in the text picture, the designated point position information of multiple characters in each text line is obtained. The designated point position information of the character represents the specific position of the character in the text image. The position information can calculate the distortion curve of each text line. The distortion curve represents the distortion of the text line. Based on the distortion of each text line, the original distorted text image can be corrected into an undistorted text image through correction mapping. Correction of the distorted text content in the text picture is realized, and the user's reading experience is improved.

For the embodiments of the electronic device, the machine-readable storage medium and the display system, since the content of the methods involved are basically similar to the foregoing method embodiments, the description is relatively simple. Can.

It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any relationship between these entities or operations. any such actual relationship or sequence exists. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device that includes a list of elements includes not only those elements, but also includes not explicitly listed or other elements inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

Each embodiment in this specification is described in a related manner, and the same and similar parts between the various embodiments may be referred to each other, and each embodiment focuses on the differences from other embodiments. Especially, for the apparatus, electronic device, machine-readable storage medium, and display system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and for related parts, please refer to some descriptions of the method embodiments.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims (14)

1. A method for rectifying a text picture, the method comprising:

acquiring a text picture to be corrected;

performing text line detection on the text picture to obtain appointed point position information of a plurality of characters in each text line of the text picture;

calculating a distortion curve of each text line according to the designated point position information of the characters in each text line;

and carrying out correction mapping on the text picture based on the distortion curve of each text line to obtain a corrected text picture.

2. The method according to claim 1, wherein the performing text line detection on the text picture to obtain the position information of the designated points of the plurality of characters in each text line of the text picture comprises:

carrying out document area detection on the text picture to obtain a detection box coordinate of a document area in the text picture;

and according to the detection frame coordinates, detecting and obtaining the appointed point position information of a plurality of characters in each text line of the document area from the document area.

3. The method according to claim 1, wherein said calculating a warping curve for each text line based on the specified point position information of the plurality of characters in each text line comprises:

constructing an initialization curve;

calculating an error value between a first vertical coordinate in the designated point position information and a second vertical coordinate on the initialization curve according to the designated point position information of the characters in each text line, and constructing an error function of each text line corresponding to the initialization curve, wherein the second vertical coordinate is a vertical coordinate on the initialization curve corresponding to the same horizontal coordinate in the designated position information;

and obtaining a distortion curve of each text line through curve fitting according to the error function of each text line corresponding to the initialization curve.

4. The method of claim 3, wherein said obtaining a warping curve for each text line by curve fitting based on an error function of each text line corresponding to the initialization curve comprises:

and carrying out minimization operation on the error function of each text line corresponding to the initialization curve, and fitting to obtain a distortion curve of each text line.

5. The method according to claim 1, wherein said performing rectification mapping on the text picture based on the distortion curve of each text line to obtain a rectified text picture comprises:

forming a three-dimensional grid by the distortion curves of the text lines to form a distortion curved surface;

and mapping the distorted curved surface to a plane picture to obtain a corrected text picture.

6. The method according to claim 1, wherein after said performing rectification mapping on the text picture based on the warping curve of each text line to obtain a rectified text picture, the method further comprises:

and recognizing each character in the corrected text picture by adopting a character recognition technology.

7. A text picture rectification apparatus, characterized in that the apparatus comprises:

the acquisition module is used for acquiring a text picture to be corrected;

the detection module is used for detecting text lines of the text picture to obtain the position information of the appointed points of a plurality of characters in each text line of the text picture;

the calculation module is used for calculating the distortion curve of each text line according to the designated point position information of the characters in each text line;

and the correction module is used for performing correction mapping on the text image based on the distortion curve of each text line to obtain a corrected text image.

8. The apparatus according to claim 7, wherein the detection module is specifically configured to:

carrying out document area detection on the text picture to obtain a detection box coordinate of a document area in the text picture;

and according to the detection frame coordinates, detecting and obtaining the appointed point position information of a plurality of characters in each text line of the document area from the document area.

9. The apparatus of claim 7, wherein the computing module is specifically configured to:

constructing an initialization curve;

calculating an error value between a first vertical coordinate in the designated point position information and a second vertical coordinate on the initialization curve according to the designated point position information of the characters in each text line, and constructing an error function of each text line corresponding to the initialization curve, wherein the second vertical coordinate is a vertical coordinate on the initialization curve corresponding to the same horizontal coordinate in the designated position information;

and obtaining a distortion curve of each text line through curve fitting according to the error function of each text line corresponding to the initialization curve.

10. The apparatus according to claim 9, wherein the computing module, when configured to obtain the warping curve of each text line by curve fitting according to the error function of each text line corresponding to the initialization curve, is specifically configured to:

and carrying out minimization operation on the error function of each text line corresponding to the initialization curve, and fitting to obtain a distortion curve of each text line.

11. The device according to claim 7, wherein the corrective module is specifically configured to:

forming a three-dimensional grid by the distortion curves of the text lines to form a distortion curved surface;

and mapping the distorted curved surface to a plane picture to obtain a corrected text picture.

12. The apparatus of claim 7, further comprising:

and the character recognition module is used for recognizing each character in the corrected text picture by adopting a character recognition technology.

13. An electronic device comprising a processor and a memory, wherein,

the memory is used for storing a computer program;

the processor, when executing the computer program stored on the memory, implementing the method steps of any of claims 1-6.

14. A machine readable storage medium, characterized in that a computer program is stored in the machine readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of claims 1-6.

Images