WO2021147221A1

WO2021147221A1 - Text recognition method and apparatus, and electronic device and storage medium

Info

Publication number: WO2021147221A1
Application number: PCT/CN2020/093605
Authority: WO
Inventors: 肖京; 何嘉欣; 刘鹏; 刘玉宇
Original assignee: 平安科技（深圳）有限公司
Priority date: 2020-01-22
Filing date: 2020-05-30
Publication date: 2021-07-29
Also published as: CN111325104A; CN111325104B

Abstract

The present application relates to artificial intelligence, and in particular to the field of image processing. Disclosed is a text recognition method. The method comprises: after receiving an instruction which carries an image to be recognized and which is sent by a user, performing text recognition on the image to be recognized, so as to obtain a first target recognition result; generating, according to the first target recognition result, a picture to be verified, calculating the similarity between the picture to be recognized and a corresponding target text box, recognizing an abnormal text box in the first target recognition result according to the similarity, and performing abnormality processing; updating the first target recognition result on the basis of an abnormality processing result to obtain a second target recognition result; and feeding back the second target recognition result to the user. Further disclosed are a text recognition apparatus, an electronic device and a computer storage medium. The present application can improve the accuracy of text recognition.

Description

Text recognition method, device, electronic equipment and storage medium

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office, the application number is 202010073495.6, and the invention title is "text recognition method, device and storage medium" on January 22, 2020, the entire content of which is incorporated into the application by reference .

Technical field

This application relates to the field of artificial intelligence, and in particular to a text recognition method, device, electronic equipment, and computer-readable storage medium.

Background technique

Nowadays, dedicated OCR recognition already has a set of mature algorithms, which are respectively responsible for target file detection, field detection and field recognition. This process is end-to-end, and the results will be directly output to the user.

The basic process of the existing general OCR recognition is to first detect the area where the text in the picture is located, draw the circumscribed rectangular frame of each area, and then perform a basic two-dimensional rotation correction on each rectangular frame, and then input the cut block for recognition Module to obtain all the text content of the entire picture. Although this process can correct the inclination of the target in a two-dimensional plane, the inventor realizes that in actual image recognition scenarios, there are often cases where the recognition target and the original picture are not coplanar. In this case, the image recognition result will be far from the correct result.

Therefore, it is urgent to provide a method that can accurately recognize text from pictures.

technical problem

Based on this, it is necessary to provide a text recognition method, device, electronic device, and computer-readable storage medium that can improve the accuracy of text recognition in response to the above technical problems.

Technical solutions

A text recognition method, the method includes:

Receiving step: receiving a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized;

Recognition step: perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized, and the first target recognition result includes multiple target text boxes and the multiple targets The first target text information corresponding to the text box;

Analysis step: generate pictures to be verified corresponding to the plurality of target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and obtain the results from all of them according to the model output results. An abnormal text box is recognized in the first target recognition result;

Update step: send the abnormal text box to a preset terminal, and receive the second target text information of the abnormal text box fed back by the preset terminal, and update all information based on the second target text information of the abnormal text box. The first target recognition result is described, and the second target recognition result is generated;

The first feedback step: feedback the second target recognition result to the user.

A text recognition device, which includes:

A receiving module, configured to receive a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized;

The recognition module is configured to perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized. The first target recognition result includes multiple target text boxes and the multiple First target text information corresponding to each target text box;

The analysis module is configured to generate pictures to be verified corresponding to the multiple target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and output the results according to the model Identifying an abnormal text box from the first target recognition result;

An update module, configured to send the abnormal text box to a preset terminal, and receive the second target text information of the abnormal text box fed back by the preset terminal, based on the second target text information of the abnormal text box Update the first target recognition result to generate a second target recognition result;

The first feedback module is configured to feed back the second target recognition result to the user.

An electronic device comprising: a memory and a processor, the memory storing a text recognition program that can be run on the processor, and the text recognition program can be executed by the processor to implement the following steps:

A computer-readable storage medium, the computer-readable storage medium includes a text recognition program, and when the text recognition program is executed by a processor, the following steps can be implemented:

The first feedback step: feedback the second target recognition result to the user

Beneficial effect

The above text recognition method, device, electronic equipment, and computer-readable storage medium, after receiving the instruction from the user to carry the image to be recognized, perform text recognition on the image to be recognized to obtain the first target recognition result, and generate according to the first target recognition result For the image to be verified, calculate the similarity between the image to be recognized and the corresponding target text box, identify the abnormal text box in the first target recognition result based on the similarity and perform exception processing, and update the first target recognition result based on the abnormal processing result to obtain the second target The recognition result, the second target recognition result is fed back to the user. By adding a verification mechanism after the general OCR recognition process, the accuracy of the output of the recognition results is improved, and the user experience is improved; by random perspective transformation of the image to be recognized, the accuracy rate is selected from the recognition results corresponding to the results of multiple perspective transformations The highest text information is used as the first target text information of the target text box, which improves the accuracy of text recognition; before the image to be recognized is recognized, the image to be recognized is also subjected to distortion correction, which lays the foundation for accurate text recognition.

Description of the drawings

Figure 1 is a flowchart of a preferred embodiment of the text recognition method of this application;

FIG. 2 is a schematic diagram of a preferred embodiment of the electronic device of this application;

FIG. 3 is a schematic diagram of modules of a preferred embodiment of the text recognition device of this application.

Embodiments of the present invention

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 text recognition method. The method can be executed by a device, and the device can be implemented by software and/or hardware.

Referring to FIG. 1, it is a flowchart of a preferred embodiment of the text recognition method of this application.

In a preferred embodiment of the text recognition method of the present application, the text recognition method only includes: step S1-step S5.

Step S1: Receive a text recognition instruction sent by a user, where the text recognition instruction includes an image to be recognized.

In the following, the electronic device is used as the execution subject to describe each embodiment of the present application.

The user selects the image to be recognized through the APP on the client, and sends a text recognition instruction based on the selected image to be recognized. After receiving the instruction issued by the client, the electronic device performs a text recognition operation on the image to be recognized carried in the instruction.

Step S2: Perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized. The first target recognition result includes multiple target text boxes and the multiple targets The first target text information corresponding to the text box.

For example, the pre-trained OCR recognition model is used to perform OCR recognition on the image to be recognized, and the recognition result output by the model is used as the first target recognition result.

Step S3, generating pictures to be verified corresponding to the multiple target text boxes based on the first target recognition result, inputting the pictures to be verified and the target text boxes into a preset analysis model, and outputting the results from all An abnormal text box is identified in the first target recognition result.

In order to improve the accuracy of text recognition, after the first target recognition result is obtained by performing OCR recognition on the picture to be recognized, the accuracy of the first target recognition result is verified. Therefore, it is necessary to perform abnormal text box recognition on the first target recognition result. In this embodiment, the generating the to-be-verified pictures corresponding to the multiple target text boxes based on the first target recognition result includes:

a1. Read the length and width information of one of the target text boxes in the first recognition result, and create a new background picture consistent with the length and width information of the target text box; and

a2. Obtain the first target text information corresponding to the target text box in the first recognition result, and place the first target text information corresponding to the target text box in the background picture in a preset format to generate all The image to be verified corresponding to the target text box.

For example, take a target text box P as an example, read the length and width information of the target text box P, use the length and width information to determine a pure color background picture P1 with a random color (light color is best, for example, white), and then obtain The first target text information PT corresponding to the target text box P, the first target text information PT is formatted to generate the first target text information PT1 of Song Ti, and the first target text information PT1 of Song Ti is centered on the pure color background picture P1 , To obtain a picture P2 to be verified with black characters on a white background and corresponding to the target text box P.

In other embodiments, the placing the first target text information corresponding to the target text box in the solid-color background picture in a preset format further includes:

b1. Perform random format adjustment on the first target text information corresponding to the target text box; and

b2. Place the first target text information adjusted in a random format in the background picture.

For example, modify the font color, font size, font, angle, etc. Through the above random adjustments, the randomness of the generated images to be verified is increased, laying the foundation for subsequent accurate verification.

In this embodiment, after the corresponding picture to be verified is generated based on each target text box, each target text box and its corresponding picture to be verified are compared and analyzed, and the consistency of the target text box and its corresponding picture to be verified is carried out. Analysis can determine the abnormal text box in the first target recognition result.

In this embodiment, the preset analysis model is a convolutional neural network, and preferably, the preset analysis model is resnet50. The preset analysis model is used to extract features from the target text box and its corresponding image to be verified. Pre-train a convolutional neural network for feature extraction, use the trained neural network to extract the features of the text box and the image to be verified, and calculate the similarity between the text box and the image to be verified to determine whether the contents of the two images are the same , And perform exception handling on the text boxes judged to be inconsistent.

The analysis model includes a batch input layer, a feature extraction layer, an L2 normalization layer, and a loss function. The loss function includes but not limited to Softmax loss, Center Either loss or Triplet loss. Different loss functions have different requirements for training data.

Take the triplet loss function as an example. In this embodiment, in order to adapt to the triplet loss function, it is necessary to select anchor point samples, positive samples, and negative samples in the batch. In the OCR business, the anchor point sample refers to the field interception map of the original image. , The positive sample refers to the graph generated according to the field content, and the negative sample refers to the graph produced by replacing the field content. Among them, for the selection of negative samples, each replaced field can be replaced according to the order of the Chinese character table.

In other embodiments, in order to adapt to the central loss function, it is necessary to select n pictures generated according to different sizes, angles, and colors in each type of sample, and n copied field cropping pictures. In the process of generating sample pictures, set the font size interval, angle interval, and color interval, and randomly select values in the interval to obtain transformation parameters to generate the picture.

During training, the ROC curve is used to calculate the threshold that maximizes the accuracy of the model. Through training, the hope to achieve the effect is: in the feature space, the distance between the image features of the same content is getting closer, and the distance between the image features of other different content is getting farther and farther.

In this embodiment, the inputting the image to be verified and the target text box into a preset analysis model, and identifying an abnormal text box from the first target recognition result according to the model output result includes:

c1. Determine the feature vectors of the target text box and the picture to be verified respectively according to the output result of the model;

c2. Calculate the similarity between the target text box and the feature vector of the picture to be verified by using a preset similarity algorithm; and

c3. When the similarity is less than a preset similarity threshold, determine that the target text box is an abnormal text box.

For example, the aforementioned preset similarity algorithm includes but is not limited to any one of Euclidean distance algorithm and cosine similarity algorithm.

It is understandable that the features extracted from the target text box can better reflect the original features of the region in the image to be recognized, and the features extracted from the corresponding image to be verified can better reflect the features of the first target text information. Calculate the similarity between the two to judge the consistency between the two. The higher the similarity, the higher the likelihood of the two being consistent, that is, the higher the accuracy of the recognition result. Conversely, the lower the similarity, the lower the likelihood of the two being consistent, that is, the lower the accuracy of the recognition result. By setting a similarity threshold, the target text boxes with similarity greater than or equal to the similarity threshold are regarded as normal text boxes, and the target text boxes with similarity less than the similarity threshold are regarded as abnormal text boxes.

Step S4, the abnormal text box is sent to a preset terminal, and the second target text information of the abnormal text box fed back by the preset terminal is received, and the second target text information of the abnormal text box is updated based on the second target text information of the abnormal text box. According to the first target recognition result, the second target recognition result is generated.

After determining the abnormal text box in the first target recognition result, the abnormal text box needs to be processed. In this embodiment, the aforementioned preset terminal is a terminal used by crowdsourced personnel. Send the abnormal text box to the crowdsourcing personnel, artificially identify the second target text information corresponding to the abnormal text box, and return the second target text information corresponding to the abnormal text box to the electronic device. The electronic device updates the first target text information corresponding to the abnormal text box in the first target recognition result based on the received second target text information corresponding to the abnormal text box to obtain the second target recognition result.

Step S5, feeding back the second target recognition result to the user.

After determining the second target recognition result of the image to be recognized, the second target recognition result is displayed to the user through the client.

In other embodiments, the text recognition method includes: steps S1-step S3 and step S6.

Step S6: When there is no abnormal text box in the first recognition result, the first recognition result is fed back to the user.

When it is determined that there is no abnormal text box in the first target recognition result, the first target recognition result is directly used as the final recognition result, and the obtained final recognition result is displayed to the user through the client.

In other embodiments, the performing text recognition on the image to be recognized based on preset recognition rules to obtain the first target recognition result of the image to be recognized includes:

d1. Identify the field area of the image to be recognized, and determine a plurality of first text boxes of the image to be recognized;

d2. Perform multiple random perspective transformations on each of the first text boxes to obtain multiple second text boxes corresponding to each of the first text boxes;

d3. Input the multiple second text boxes corresponding to each of the first text boxes into a preset recognition model to obtain first recognition results of the multiple second text boxes corresponding to each of the first text boxes;

d4. Filter out the first text box from the plurality of second text boxes corresponding to each of the first text boxes based on the first recognition results of the plurality of second text boxes corresponding to each of the first text boxes The corresponding target text box; and

d5. Determine the first target recognition result of the image to be recognized according to the first recognition result of the target text box corresponding to each of the first text boxes.

Firstly, detect the position of the text field in the image to be recognized, and determine the circumscribed rectangular box containing the position of the text field, that is, the first text box.

Then, multiple random perspective transformations are performed on each first text box to obtain multiple corresponding second text boxes. For example, perform 5 random perspective transformations on each first text box to obtain 5 second text boxes corresponding to one first text box. The second text box includes the first text box.

Then, the OCR recognition model is used to identify the first text information and the first confidence level corresponding to the five second text boxes. And filter out the second text box with the highest first confidence as the target text box corresponding to the first text box.

Finally, the first target text information of the first text box is determined according to the first text information of the target text box, and the first target text information of each first text box is summarized to obtain the first target recognition result.

It is understandable that the image to be recognized may be captured by the user in real time, and when the user uses the camera to capture the image to be recognized, the image may be distorted due to the characteristics of the camera itself. Therefore, in order to further improve the accuracy of recognition, in other embodiments, before the step S2, the method further includes:

Distortion correction is performed on the image to be recognized based on a preset distortion correction rule to obtain the image to be recognized after distortion correction.

In this embodiment, the performing distortion correction on the image to be recognized based on a preset distortion correction rule to obtain the image to be recognized after distortion correction includes:

e1. Obtain the pixel corners of the image to be recognized, and calculate the coordinates of the pixel corners on the undistorted image;

e2, calculating a perspective transformation matrix according to the coordinates of the pixel corners on the undistorted image; and

e3. Perform distortion correction on the image to be recognized according to the perspective transformation matrix, and generate a distortion-corrected image to be recognized.

In this embodiment, the coordinates of each pixel corner on the undistorted image are obtained by performing distortion correction on the pixel corners on the original image to be recognized with distortion, where the pixel corners can be the distorted image to be recognized If the image to be recognized is a quadrilateral, it is the four vertices of the quadrilateral. When calculating the perspective transformation matrix, at least the corresponding coordinates of four pixel points are needed to solve the solution. Therefore, when obtaining the pixel corner points on the image to be recognized with distortion, at least the coordinates of the four pixel corner points need to be obtained. Take the two-dimensional code image as an example, you can first obtain the coordinates of the four pixel corners of the two-dimensional code area in the image from the original distorted image, that is, the coordinates of the four vertices of the two-dimensional code, and then use the previous The calibrated distortion parameters calculate the coordinates of the four corner points on the undistorted image: [x, y] = K[u, v], where [x, y] are the pixel corner coordinates on the original distorted image, [u, v] is the pixel corner coordinates on the undistorted image, and K is the distortion parameter.

After the perspective transformation matrix is solved, the image to be recognized can be subjected to distortion correction to obtain the image to be recognized after the distortion correction, and then the subsequent recognition, verification, update and feedback operations are performed.

Because the coordinates of pixel corners on the undistorted image calculated by distortion correction are not one-to-one mapping, it is possible that the coordinates on the undistorted image calculated for the pixel corners on the original distorted image are not unique, in order to find The optimal coordinates of the pixel corners on the undistorted image.

In other embodiments, the calculating the coordinates of the pixel corners on the undistorted image includes:

f1. Determine a target pixel on the undistorted image, and the coordinates of the target pixel are the same as the coordinates of the pixel corners on the image to be recognized;

f2. Determine a pixel in a circular area with the target pixel as the center and a preset neighborhood radius as the radius as the neighborhood pixel;

f3. Traverse each neighborhood pixel of the target pixel on the undistorted image, and calculate the coordinates of each neighborhood pixel on the image to be recognized; and

f4. Determine the coordinates of the pixel corners on the undistorted image according to the coordinates of the respective neighborhood pixel points on the image to be recognized.

For example, the distance between each neighborhood pixel and the pixel corner can be calculated according to the coordinates of each neighborhood pixel on the original distorted image to be recognized, and then the coordinates corresponding to the shortest distance can be determined as the pixel corner is undistorted The coordinates on the image. When determining the coordinates of each pixel corner on the original distorted image to be recognized in the undistorted image, the neighborhood radius can be flexibly set according to the degree of distortion of the original distorted image to be recognized. When the degree of distortion is small, the neighborhood radius It can be set smaller, so that there are fewer pixels in the neighborhood that need to be traversed, which can reduce the amount of calculation. When the degree of distortion is large, the radius of the neighborhood can be set larger, so that the optimal pixel can be found.

In the text recognition method proposed in the above embodiment, after receiving an instruction from a user to carry an image to be recognized, text recognition is performed on the image to be recognized to obtain a first target recognition result, the image to be verified is generated according to the first target recognition result, and the to-be-recognized image is calculated The similarity between the picture and the corresponding target text box, the abnormal text box in the first target recognition result is identified according to the similarity and the exception processing is performed, and the first target recognition result is updated based on the abnormal processing result to obtain the second target recognition result, and the second target The recognition result is fed back to the user. By adding a verification mechanism after the general OCR recognition process, the accuracy of the output of the recognition results is improved, and the user experience is improved; by random perspective transformation of the image to be recognized, the accuracy rate is selected from the recognition results corresponding to the results of multiple perspective transformations The highest text information is used as the first target text information of the target text box, which improves the accuracy of text recognition; before the image to be recognized is recognized, the image to be recognized is also subjected to distortion correction, which lays the foundation for accurate text recognition.

This application also proposes an electronic device. Refer to FIG. 2, which is a schematic diagram of a preferred embodiment of the electronic device of this application.

In this embodiment, the electronic device 1 may be a terminal device with data processing functions such as a server, a smart phone, a tablet computer, a portable computer, a desktop computer, etc. The server may be a rack server, a blade server, or a tower. Server or rack server.

The electronic device 1 includes a memory 11, a processor 12 and a network interface 13.

The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, and the like. The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a 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 hard disk, a smart memory card (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, for example, a text recognition program 10, etc., but also to temporarily store data that has been output or will be output.

The processor 12 may be a central processing unit (Central Processing Unit) in some embodiments. Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips are used to run program codes or processing data stored in the memory 11, for example, the text recognition program 10, etc.

The network interface 13 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface), which is usually used to establish a communication connection between the electronic device 1 and other electronic devices, for example, a client (not shown in the figure). ). The components 11-13 of the electronic device 1 communicate with each other via a communication bus.

FIG. 2 only shows the electronic device 1 with components 11-13. Those skilled in the art can understand that the structure shown in FIG. 2 does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure. Multiple components, or a combination of certain components, or different component arrangements.

Optionally, the electronic device 1 may further include a user interface. The user interface may include a display (Display) and an input unit such as a keyboard (Keyboard). The optional user interface may also include a standard wired interface and 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 organic light-emitting diode (Organic Light-Emitting Diode, OLED) touch device, etc. Among them, the display may also be 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.

In the embodiment of the electronic device 1 shown in FIG. 2, the memory 11 as a computer storage medium stores the program code of the text recognition program 10, and when the processor 12 executes the program code of the text recognition program 10, the following steps are implemented:

Receiving step: receiving a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized.

The user selects the image to be recognized through the APP on the client, and sends a text recognition instruction based on the selected image to be recognized. After receiving the instruction sent by the client, the electronic device 1 performs a text recognition operation on the image to be recognized carried in the instruction.

Recognition step: perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized, and the first target recognition result includes multiple target text boxes and the multiple targets The first target text information corresponding to the text box.

Analysis step: generate pictures to be verified corresponding to the plurality of target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and obtain the results from all of them according to the model output results. An abnormal text box is identified in the first target recognition result.

For example, take a target text box P as an example, read the length and width information of the target text box P, use the length and width information to determine a pure color background picture P1 with a random color (light color is best, for example, white), and then obtain The first target text information PT corresponding to the target text box P, the first target text information PT is formatted to generate the first target text information PT1 of Song Ti, and the first target text information PT1 of Song Ti is centered on the pure color background picture P1 , A picture P2 to be verified with black characters on a white background corresponding to the target text box P is obtained.

After determining the abnormal text box in the first target recognition result, the abnormal text box needs to be processed. In this embodiment, the aforementioned preset terminal is a terminal used by crowdsourced personnel. The abnormal text box is sent to the crowdsourcing personnel, the second target text information corresponding to the abnormal text box is manually recognized, and the second target text information corresponding to the abnormal text box is returned to the electronic device 1. The electronic device 1 updates the first target text information corresponding to the abnormal text box in the first target recognition result based on the received second target text information corresponding to the abnormal text box to obtain the second target recognition result.

Feedback step: feedback the second target recognition result to the user.

In other embodiments, when there is no abnormal text box in the first recognition result, the first recognition result is fed back to the user. When it is determined that there is no abnormal text box in the first target recognition result, the first target recognition result is directly used as the final recognition result, and the obtained final recognition result is displayed to the user through the client.

It is understandable that the image to be recognized may be captured by the user in real time, and when the user uses the camera to capture the image to be recognized, the image may be distorted due to the characteristics of the camera itself. Therefore, in order to further improve the accuracy of recognition, in other embodiments, when the processor 12 executes the text recognition program 10, before the recognition step, the following steps are further implemented:

This application also proposes a text recognition device.

Referring to FIG. 3, it is a schematic diagram of modules of a preferred embodiment of the text recognition device of this application.

The text recognition apparatus 2 in this embodiment may include: module 210-module 250 according to the realized functions. The module can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of the electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

In an embodiment of the text recognition device 2 of the present application, the functions of each module/unit are as follows:

The receiving module 210 is configured to receive a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized;

The recognition module 220 is configured to perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized. The first target recognition result includes a plurality of target text boxes and the The first target text information corresponding to the multiple target text boxes;

The analysis module 230 is configured to generate pictures to be verified corresponding to the multiple target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and output according to the model As a result, an abnormal text box is identified from the first target recognition result;

The update module 240 is configured to send the abnormal text box to a preset terminal, and receive the second target text information of the abnormal text box fed back by the preset terminal, based on the second target text of the abnormal text box Information updates the first target recognition result, and generates a second target recognition result;

The feedback module 250 is configured to feed back the second target recognition result to the user.

The functions or operation steps implemented by the modules 210-250 are similar to the above, and will not be described in detail here.

In addition, the embodiment of the present application also proposes a computer-readable storage medium. The computer-readable storage medium may be non-volatile or volatile. The computer-readable storage medium includes a text recognition program 10, which implements any step of the text recognition method when the text recognition program 10 is executed by a processor. The specific implementation of the computer-readable storage medium of the present application is substantially the same as the foregoing method embodiment, and will not be repeated here.

The serial numbers of the foregoing embodiments of the present application are for description only, and do not represent the advantages and disadvantages of the embodiments.

It should be noted that in this article, the terms "include", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements not only includes those elements, It also includes other elements not explicitly listed, or elements inherent to the process, device, article, or method. If there are no more restrictions, the element defined by the sentence "including a..." does not exclude the existence of other identical elements in the process, device, article, or method that includes the element.

Through the description of the above implementation manners, those skilled in the art can clearly understand that the above-mentioned embodiment method can be implemented by means of software plus the necessary general hardware platform, of course, it can also be implemented by hardware, but in many cases the former is better.的实施方式。 Based on this understanding, the technical solution of this application essentially or the part that contributes to the existing technology can be embodied in the form of a software product, and the computer software product is stored in a storage medium (such as ROM/RAM) as described above. , Magnetic disks, optical disks), including several instructions to make a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) execute the methods described in the various embodiments of this application.

The above are only the preferred embodiments of the application, and do not limit the scope of the patent for this application. Any equivalent structure or equivalent process transformation made using the content of the description and drawings of the application, or directly or indirectly applied to other related technical fields , The same reason is included in the scope of patent protection of this application.

Claims

A text recognition method suitable for electronic equipment, wherein the method includes:

Receiving step: receiving a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized;

Recognition step: perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized, and the first target recognition result includes multiple target text boxes and the multiple targets The first target text information corresponding to the text box;

Analysis step: generate pictures to be verified corresponding to the plurality of target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and obtain the results from all of them according to the model output results. An abnormal text box is recognized in the first target recognition result;

Update step: send the abnormal text box to a preset terminal, and receive the second target text information of the abnormal text box fed back by the preset terminal, and update all information based on the second target text information of the abnormal text box. The first target recognition result is described, and the second target recognition result is generated;

The first feedback step: feedback the second target recognition result to the user.
The text recognition method according to claim 1, wherein the text recognition method further comprises:

The second feedback step: when there is no abnormal text box in the first recognition result, the first recognition result is fed back to the user.
The text recognition method according to claim 1, wherein said generating the to-be-verified pictures corresponding to the multiple target text boxes based on the first target recognition result comprises:

Read the length and width information of one of the target text boxes in the first recognition result, and create a new background picture consistent with the length and width information of the target text box; and

Acquire the first target text information corresponding to the target text box in the first recognition result, place the first target text information corresponding to the target text box in the background picture in a preset format, and generate the target The image to be verified corresponding to the text box.
The text recognition method according to claim 3, wherein the placing the first target text information corresponding to the target text box in the background picture in a preset format further comprises:

Performing random format adjustment on the first target text information corresponding to the target text box; and

The first target text information adjusted in a random format is placed in the background picture.
The text recognition method according to any one of claims 1 to 4, wherein said inputting said image to be verified and said target text box into a preset analysis model, and extracting from said first target according to the model output result An abnormal text box is identified in the recognition result, including:

Respectively determine the feature vectors of the target text box and the picture to be verified according to the output result of the model;

Using a preset similarity algorithm to calculate the similarity between the target text box and the feature vector of the picture to be verified; and

When the similarity is less than the preset similarity threshold, it is determined that the target text box is an abnormal text box.
The text recognition method according to claim 1, wherein the preset similarity algorithm includes but is not limited to any one of Euclidean distance algorithm and cosine similarity algorithm.
The text recognition method according to claim 1, wherein the performing text recognition on the image to be recognized based on preset recognition rules to obtain the first target recognition result of the image to be recognized comprises:

Identifying the field area of the image to be recognized, and determining a plurality of first text boxes of the image to be recognized;

Performing multiple random perspective transformations on each of the first text boxes to obtain multiple second text boxes corresponding to each of the first text boxes;

Inputting multiple second text boxes corresponding to each of the first text boxes into a preset recognition model to obtain first recognition results of the multiple second text boxes corresponding to each of the first text boxes;

Based on the first recognition results of the plurality of second text boxes corresponding to each of the first text boxes, filter out the plurality of second text boxes corresponding to each of the first text boxes to filter out those corresponding to the first text box Target text box; and

The first target recognition result of the image to be recognized is determined according to the first recognition result of the target text box corresponding to each of the first text boxes.
The text recognition method according to claim 1, wherein, before the recognition step, the method further comprises:

Distortion correction is performed on the image to be recognized based on a preset distortion correction rule to obtain the image to be recognized after distortion correction.
An electronic device, wherein the device includes a memory and a processor, the memory stores a text recognition program that can run on the processor, and when the text recognition program is executed by the processor, the following steps can be implemented :

Receiving step: receiving a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized;

Recognition step: perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized, and the first target recognition result includes multiple target text boxes and the multiple targets The first target text information corresponding to the text box;

Analysis step: generate pictures to be verified corresponding to the plurality of target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and obtain the results from all of them according to the model output results. An abnormal text box is recognized in the first target recognition result;

Update step: send the abnormal text box to a preset terminal, and receive the second target text information of the abnormal text box fed back by the preset terminal, and update all information based on the second target text information of the abnormal text box. The first target recognition result is described, and the second target recognition result is generated;

The first feedback step: feedback the second target recognition result to the user.
The electronic device according to claim 9, wherein the following steps are further implemented when the text recognition program is executed by the processor:

The second feedback step: when there is no abnormal text box in the first recognition result, the first recognition result is fed back to the user.
9. The electronic device according to claim 9, wherein said generating the to-be-verified pictures corresponding to said multiple target text boxes based on said first target recognition result comprises:

Read the length and width information of one of the target text boxes in the first recognition result, and create a new background picture consistent with the length and width information of the target text box; and

Acquire the first target text information corresponding to the target text box in the first recognition result, place the first target text information corresponding to the target text box in the background picture in a preset format, and generate the target The image to be verified corresponding to the text box.
11. The electronic device according to claim 11, wherein the placing the first target text information corresponding to the target text box in the background picture in a preset format further comprises:

Performing random format adjustment on the first target text information corresponding to the target text box; and

The first target text information adjusted in a random format is placed in the background picture.
The electronic device according to any one of claims 9 to 12, wherein the inputting the picture to be verified and the target text box into a preset analysis model, and identifying from the first target according to the model output result Unusual text boxes were identified in the results, including:

Respectively determine the feature vectors of the target text box and the picture to be verified according to the output result of the model;

Using a preset similarity algorithm to calculate the similarity between the target text box and the feature vector of the picture to be verified; and

When the similarity is less than the preset similarity threshold, it is determined that the target text box is an abnormal text box.
The electronic device according to claim 9, wherein the preset similarity algorithm includes but is not limited to any one of Euclidean distance algorithm and cosine similarity algorithm.
9. The electronic device according to claim 9, wherein the performing text recognition on the image to be recognized based on a preset recognition rule to obtain the first target recognition result of the image to be recognized comprises:

Identifying the field area of the image to be recognized, and determining a plurality of first text boxes of the image to be recognized;

Performing multiple random perspective transformations on each of the first text boxes to obtain multiple second text boxes corresponding to each of the first text boxes;

Inputting multiple second text boxes corresponding to each of the first text boxes into a preset recognition model to obtain first recognition results of the multiple second text boxes corresponding to each of the first text boxes;

Based on the first recognition results of the plurality of second text boxes corresponding to each of the first text boxes, filter out the plurality of second text boxes corresponding to each of the first text boxes to filter out the first text box Target text box; and

The first target recognition result of the image to be recognized is determined according to the first recognition result of the target text box corresponding to each of the first text boxes.
9. The electronic device according to claim 9, wherein, before the recognition step, the following steps are further implemented when the text recognition program is executed by the processor:

Distortion correction is performed on the image to be recognized based on a preset distortion correction rule to obtain the image to be recognized after distortion correction.
A text recognition device, wherein the text recognition device includes:

A receiving module, configured to receive a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized;

The recognition module is configured to perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized. The first target recognition result includes multiple target text boxes and the multiple First target text information corresponding to each target text box;

The analysis module is configured to generate pictures to be verified corresponding to the multiple target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and output the results according to the model Identifying an abnormal text box from the first target recognition result;

An update module, configured to send the abnormal text box to a preset terminal, and receive the second target text information of the abnormal text box fed back by the preset terminal, based on the second target text information of the abnormal text box Update the first target recognition result to generate a second target recognition result;

The first feedback module is configured to feed back the second target recognition result to the user.
A computer-readable storage medium, wherein the computer-readable storage medium includes a text recognition program, and when the text recognition program is executed by a processor, the following steps can be implemented:

Receiving step: receiving a text recognition instruction issued by a user, where the text recognition instruction includes an image to be recognized;

Recognition step: perform text recognition on the image to be recognized based on preset recognition rules to obtain a first target recognition result of the image to be recognized, and the first target recognition result includes multiple target text boxes and the multiple targets The first target text information corresponding to the text box;

Analysis step: generate pictures to be verified corresponding to the plurality of target text boxes based on the first target recognition result, input the pictures to be verified and the target text boxes into a preset analysis model, and obtain the results from all of them according to the model output results. An abnormal text box is recognized in the first target recognition result;

Update step: send the abnormal text box to a preset terminal, and receive the second target text information of the abnormal text box fed back by the preset terminal, and update all information based on the second target text information of the abnormal text box. The first target recognition result is described, and the second target recognition result is generated;

The first feedback step: feedback the second target recognition result to the user.
18. The computer-readable storage medium according to claim 18, wherein said generating the to-be-verified pictures corresponding to the plurality of target text boxes based on the first target recognition result comprises:

Read the length and width information of one of the target text boxes in the first recognition result, and create a new background picture consistent with the length and width information of the target text box; and

Acquire the first target text information corresponding to the target text box in the first recognition result, place the first target text information corresponding to the target text box in the background picture in a preset format, and generate the target The image to be verified corresponding to the text box.
18. The computer-readable storage medium according to claim 19, wherein the placing the first target text information corresponding to the target text box in the background picture in a preset format further comprises:

Performing random format adjustment on the first target text information corresponding to the target text box; and

The first target text information adjusted in a random format is placed in the background picture.