WO2020228171A1

WO2020228171A1 - Data enhancement method and device, and computer readable storage medium

Info

Publication number: WO2020228171A1
Application number: PCT/CN2019/102195
Authority: WO
Inventors: 许洋; 刘鹏; 王健宗
Original assignee: 平安科技（深圳）有限公司
Priority date: 2019-05-16
Filing date: 2019-08-23
Publication date: 2020-11-19
Also published as: CN110287988B; CN110287988A

Abstract

Disclosed is a data enhancement method. The method comprises: obtaining an original picture; performing a first processing operation on the original picture to obtain a first processed picture, and combining the first processed picture with the original picture to obtain a first augmented picture set; performing a second processing operation on the first augmented picture set to obtain a processed first augmented picture set, and combining the processed first augmented picture set with the first augmented picture set to obtain a second augmented picture set; performing a third processing operation on the second augmented picture set to obtain a processed second augmented picture set, and combining the processed second augmented picture set with the second augmented picture set to obtain a third augmented picture set; and using the third augmented picture set as an enhanced picture set. The present application also provides a data enhancement device and a computer readable storage medium. The present application improves the diversity of data and enhances the data to implement full sample training.

Description

Data enhancement method, device and computer readable storage medium

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on May 16, 2019. The application number is 201910416659.8 and the invention title is "Data Enhancement Method, Device and Computer-readable Storage Medium". The entire content is incorporated by reference. In this application.

Technical field

This application relates to the field of computer technology, and in particular to a data enhancement method, device and computer-readable storage medium.

Background technique

The OCR services so far are basically data-driven. These services rely on a large amount of data for deep learning (for example, neural network) model training and used for the detection and recognition of target fields. Most of the applied deep learning structures, especially supervised neural networks, require a lot of training data for learning. The training data needs to manually mark the position of the required field on the picture and the corresponding real text value before the detection and recognition model can be trained.

In actual business, it is often difficult to cover all scenes with manually collected pictures, and the speed of manual labeling cannot keep up with the construction period, resulting in a situation where the amount of final labeling data cannot achieve the best results for the detection and recognition model. The accuracy of the model trained in the above-mentioned case where the labeled data is very limited is usually not very high, that is, deep learning relies on a large amount of labeled data. In the absence of sufficient data support, the effect of the model trained by deep learning is average. In order to improve this situation, it is necessary to perform some processing on the existing annotation data to increase the total amount of data.

Summary of the invention

This application provides a data enhancement method, device, and computer-readable storage medium, the main purpose of which is to increase the diversity of training data, so as to be closer to the actual scene more realistically.

To achieve the above object, the present application also provides a data enhancement method, the method includes: obtaining an original picture; performing a first processing operation on the original picture to obtain a first processed picture, and combining the first processed picture with the original picture The first extended picture set is obtained by merging, wherein the first processing operation includes at least one of the following: blurring processing, noise processing; performing a second processing operation on the first extended picture set to obtain the processed first extended picture set, and Combine the processed first extended picture set and the first extended picture set to obtain a second extended picture set, where the second processing operation includes at least one of the following: a moving operation, a rotating operation, and a filling operation; performing the second extended picture set The third processing operation is to obtain the processed second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain the third extended picture set, wherein the third processing operation includes at least one of the following: Contrast processing operation, color processing operation; taking the third extended picture set as an enhanced picture set.

In addition, in order to achieve the above object, the present application also provides a data enhancement device, the device includes a memory and a processor, the memory stores a data enhancement program that can run on the processor, and the data enhancement program When executed by the processor, the following steps are implemented: obtain an original picture; perform a first processing operation on the original picture to obtain a first processed picture, and merge the first processed picture with the original picture to obtain a first extended picture set, where The first processing operation includes at least one of the following: blurring processing, noise processing; performing a second processing operation on the first extended picture set to obtain the processed first extended picture set, and combining the processed first extended picture set Merge with the first extended picture set to obtain a second extended picture set, where the second processing operation includes at least one of the following: a move operation, a rotation operation, and a filling operation; a third processing operation is performed on the second extended picture set to obtain the processed The second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain a third extended picture set, wherein the third processing operation includes at least one of the following: contrast processing operation, color processing operation; The third extended picture set is used as an enhanced picture set.

In addition, in order to achieve the above-mentioned object, the present application also provides a computer-readable storage medium with a data enhancement program stored on the computer-readable storage medium, and the data enhancement program can be executed by one or more processors to achieve The steps of the data enhancement method as described above.

This application obtains an original picture; performs a first processing operation on the original picture to obtain a first processed picture, and merges the first processed picture with the original picture to obtain a first extended picture set, wherein the first processing operation includes at least the following One: blur processing and noise processing; perform a second processing operation on the first extended picture set to obtain the processed first extended picture set, and combine the processed first extended picture set and the first extended picture set to obtain the first extended picture set 2. An extended picture set, where the second processing operation includes at least one of the following: a moving operation, a rotating operation, and a filling operation; performing a third processing operation on the second extended picture set to obtain a processed second extended picture set, and process Then the second extended picture set and the second extended picture set are combined to obtain a third extended picture set, wherein the third processing operation includes at least one of the following: a contrast processing operation, a color processing operation; and the third extended picture set is used as an enhancement The later picture sets increase the diversity of the data and enhance the data to make the training samples more sufficient.

Description of the drawings

FIG. 1 is a schematic flowchart of a data enhancement method provided by an embodiment of this application;

2 is a schematic diagram of the internal structure of a data enhancement device provided by an embodiment of the application;

FIG. 3 is a schematic diagram of modules of a data enhancement program in a data enhancement device provided by an embodiment of the application.

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

Detailed ways

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

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

In this embodiment, the data enhancement method includes:

S10. Obtain the original picture.

In this embodiment, the original pictures are from pictures in the training set. For each picture in the training set, the data addition method can be executed to obtain multiple styles of pictures corresponding to each training picture in the training set, thereby increasing The amount of training data improves the accuracy of the training model.

S11. Perform a first processing operation on the original picture to obtain a first processed picture, and merge the first processed picture with the original picture to obtain a first extended picture set.

In an embodiment, the first processing operation includes at least one of the following: blur processing and noise processing.

In a specific implementation, the performing the first processing operation on the original picture to obtain the first processed picture, and merging the first processed picture with the original picture to obtain the first extended picture set includes:

Use different window sizes to perform mean blurring on the original pictures to obtain pictures corresponding to each window size, merge the pictures corresponding to each window size with the original pictures to obtain the first merged picture set;

Perform Gaussian blur processing on each picture in the first merged picture set to obtain the Gaussian blurred picture corresponding to each picture in the first merged picture set, and combine the Gaussian blurred picture corresponding to each picture in the first merged picture set with the first merged picture Assemble and obtain a second combined picture set;

Perform Gaussian noise processing on each picture in the second merged picture set to obtain Gaussian noise processed pictures corresponding to each picture in the second merged picture set, and combine the Gaussian noise processed pictures corresponding to each picture in the second merged picture set with the second merged picture Collect and get the first extended picture set.

Among them, different window sizes are used to perform mean blurring on the original pictures to obtain pictures corresponding to each window size, and different window size values are used to obtain different pictures, thereby achieving data enhancement.

S12. Perform a second processing operation on the first extended picture set to obtain a processed first extended picture set, and combine the processed first extended picture set and the first extended picture set to obtain a second extended picture set.

In an embodiment, the second processing operation includes at least one of the following: a moving operation, a rotating operation, and a filling operation.

In a specific implementation, the second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture set and the first extended picture set are combined to obtain the second The expanded picture collection includes:

Perform mobile processing on each picture in the first extended picture set to obtain a mobile picture corresponding to each picture in the first extended picture set, and perform a mobile picture corresponding to each picture in the first extended picture set with each picture in the first extended picture set Merge to obtain the third merged picture set;

Rotate each picture in the third merged picture set by a preset angle to obtain the rotated picture corresponding to each picture in the third merged picture set, and merge the rotated picture corresponding to each picture in the third merged picture set with the third merged picture set, Obtain the fourth combined picture collection;

Filling processing is performed on the fourth combined picture set to obtain the second extended picture set.

In this embodiment, the coordinate translation operation is performed before the image is rotated, which can avoid the loss of information. After the image is rotated, many holes will appear. These holes must be filled, otherwise the picture effect will not be good, and this operation is generally called interpolation processing.

Specifically, the moving operation includes performing horizontal and vertical moving processing on the picture. The horizontal and vertical movement refers to randomly selecting a value on the x coordinate and the y coordinate (this value can be positive or negative, and the sign indicates the moving direction), and then translate the x and y coordinates respectively. This operation is divided into two types, one is that the image size does not change, so that part of the original image will not be in the image. Another is the image size change. This preserves the content of the original picture. The formula is as follows:

The matrix [x ₀ y ₀ 1] ^T is the original picture coordinates, and [x y 1] ^T is the coordinates after translation. Set different parameters dx and dy to move the image horizontally and vertically. Combine the above-mentioned enhanced data with the picture processed in this step to achieve data enhancement.

In an embodiment, the second processing operation further includes a zooming operation and a cropping operation. The second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first The extended picture set and the first extended picture set to obtain the second extended picture set further include:

Perform a random zoom operation on each picture in the fourth merged picture set to obtain the zoomed picture corresponding to each picture in the fourth merged picture set, merge the zoomed picture corresponding to each picture in the fourth merged picture set with the fourth merged picture set, Obtain the fifth combined picture collection;

Perform random cropping operations on each picture in the fifth merged picture set to obtain the cropped pictures corresponding to each picture in the fifth merged picture set, and merge the cropped pictures corresponding to each picture in the fifth merged picture set with the fifth merged picture set, Get the second extended picture set.

S13. Perform a third processing operation on the second extended picture set to obtain a processed second extended picture set, and combine the processed second extended picture set and the second extended picture set to obtain a third extended picture set.

In an embodiment, the third processing operation includes at least one of the following: a contrast processing operation and a color processing operation;

In a specific implementation, the third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain the third The expanded picture collection includes:

Perform a contrast transformation operation on each picture in the second extended picture set to obtain a contrast-transformed picture corresponding to each picture in the second extended picture set, and convert the contrast-transformed picture corresponding to each picture in the second extended picture set to each The pictures are merged to obtain the sixth merged picture set;

Perform a color transformation operation on each picture in the sixth merged picture set to obtain the color-transformed picture corresponding to each picture in the sixth merged picture set. The color-transformed picture corresponding to each picture in the sixth merged picture set is compared with the sixth picture. The combined picture sets are merged to obtain the third extended picture set.

S14. Use the third extended picture set as an enhanced picture set.

This application obtains an original picture; performs a first processing operation on the original picture to obtain a first processed picture, and merges the first processed picture with the original picture to obtain a first extended picture set, wherein the first processing operation includes at least the following One: blur processing and noise processing; perform a second processing operation on the first extended picture set to obtain the processed first extended picture set, and combine the processed first extended picture set and the first extended picture set to obtain the first extended picture set 2. An extended picture set, where the second processing operation includes at least one of the following: a moving operation, a rotating operation, and a filling operation; performing a third processing operation on the second extended picture set to obtain a processed second extended picture set, and process Then the second extended picture set and the second extended picture set are combined to obtain a third extended picture set, wherein the third processing operation includes at least one of the following: a contrast processing operation, a color processing operation; and the third extended picture set is used as an enhancement After the picture collection. In this case, eight data enhancement methods including blur, Gaussian noise, horizontal and vertical movement, rotating picture, random zoom, contrast transformation, color transformation, and random cropping can increase the diversity of data to a certain extent, and mix them. (For example, randomly select any one or several methods, stack several methods in different order, etc.) It can generate a variety of outputs. , To enhance the data, making the training samples more sufficient.

The application also provides a data enhancement device. Referring to FIG. 2, it is a schematic diagram of the internal structure of a data enhancement device provided by an embodiment of this application.

In this embodiment, the data enhancement device 1 may be a personal computer (Personal Computer, PC), or a terminal device such as a smart phone, a tablet computer, or a portable computer. The data enhancement device 1 at least includes a memory 11, a processor 12, a communication bus 13, and a network interface 14.

Wherein, 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, etc. The memory 11 may be an internal storage unit of the data enhancement device 1 in some embodiments, for example, the hard disk of the data enhancement device 1. In other embodiments, the memory 11 may also be an external storage device of the data enhancement device 1, such as a plug-in hard disk, a smart media card (SMC), and a secure digital (Secure Digital, SD card, Flash Card, etc. Further, the memory 11 may also include both an internal storage unit of the data enhancement apparatus 1 and an external storage device. The memory 11 can be used not only to store application software and various types of data installed in the data enhancement device 1, such as the code of the data enhancement program 01, etc., but also to temporarily store data that has been output or will be output.

The processor 12 may be a central processing unit (CPU), controller, microcontroller, microprocessor or other data processing chip in some embodiments, and is used to run the program code or processing stored in the memory 11 Data, such as executing data enhancement program 01, etc.

The communication bus 13 is used to realize the connection and communication between these components.

The network interface 14 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface), and is usually used to establish a communication connection between the device 1 and other electronic devices.

Optionally, the device 1 may also 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 liquid crystal display, an organic light-emitting diode (OLED) touch device, and the like. Among them, the display can also be called a display screen or a display unit as appropriate, and is used to display the information processed in the data enhancement device 1 and to display a visualized user interface.

Figure 2 only shows the data enhancement device 1 with components 11-14 and the data enhancement program 01. Those skilled in the art can understand that the structure shown in Figure 1 does not constitute a limitation on the data enhancement device 1, and may include Fewer or more components than shown, or some combination of components, or different component arrangement.

In the embodiment of the device 1 shown in FIG. 2, the data enhancement program 01 is stored in the memory 11; when the processor 12 executes the data enhancement program 01 stored in the memory 11, the following steps are implemented:

Get the original picture.

Perform a first processing operation on the original picture to obtain a first processed picture, and merge the first processed picture with the original picture to obtain a first extended picture set.

Using different window sizes to perform mean blurring on the original pictures to obtain pictures corresponding to each window size, and merge the pictures corresponding to each window size with the original pictures to obtain the first merged picture set;

A second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture set and the first extended picture set are combined to obtain the second extended picture set.

Specifically, the moving operation includes performing horizontal and vertical moving processing on the picture. The horizontal and vertical movement refers to randomly selecting a value on the x coordinate and the y coordinate (this value can be positive or negative, and the sign indicates the moving direction), and then respectively shifting the x coordinate and y coordinate. This operation is divided into two types, one is that the image size does not change, so that part of the original image will not be in the image. Another is the image size change. This preserves the content of the original picture. The formula is as follows:

A third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain the third extended picture set.

The third extended picture set is used as an enhanced picture set.

Optionally, in other embodiments, the data enhancement program may also be divided into one or more modules, and the one or more modules are stored in the memory 11 and are executed by one or more processors (in this embodiment, processing The module 12) is executed to complete the application. The module referred to in the application refers to a series of computer program instruction segments capable of completing specific functions, and is used to describe the execution process of the data enhancement program in the data enhancement device.

For example, referring to FIG. 3, which is a schematic diagram of the program modules of the data enhancement program in an embodiment of the data enhancement device of this application. In this embodiment, the data enhancement program can be divided into an acquisition module 10, a processing module 20, and a determination module 30. , Exemplarily:

The obtaining module 10 obtains the original picture;

The processing module 20 performs a first processing operation on the original picture to obtain a first processed picture, and merges the first processed picture with the original picture to obtain a first extended picture set, wherein the first processing operation includes at least one of the following: Blur processing, noise processing;

The processing module 20 performs a second processing operation on the first extended picture set to obtain the processed first extended picture set, and combines the processed first extended picture set and the first extended picture set to obtain a second extended picture set , Wherein the second processing operation includes at least one of the following: a moving operation, a rotating operation, and a filling operation;

The processing module 20 performs a third processing operation on the second extended picture set to obtain the processed second extended picture set, and combines the processed second extended picture set and the second extended picture set to obtain the third extended picture set , Wherein the third processing operation includes at least one of the following: a contrast processing operation and a color processing operation;

The determining module 30 uses the third extended picture set as the enhanced picture set.

The functions or operation steps implemented by the program modules such as the acquisition module 10, the processing module 20, and the determination module 30 when executed are substantially the same as those in the foregoing embodiment, and will not be repeated here.

In addition, an embodiment of the present application also proposes a computer-readable storage medium having a data enhancement program stored on the computer-readable storage medium, and the data enhancement program can be executed by one or more processors to implement the following operations:

Get the original picture;

Perform a first processing operation on the original picture to obtain a first processed picture, and merge the first processed picture with the original picture to obtain a first extended picture set, wherein the first processing operation includes at least one of the following: blur processing, Noise treatment

Perform a second processing operation on the first extended picture set to obtain the processed first extended picture set, and obtain the second extended picture set by combining the processed first extended picture set and the first extended picture set, wherein the second processing Operations include at least one of the following: moving operations, rotating operations, and filling operations;

The third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain the third extended picture set, where the third processing The operations include at least one of the following: contrast processing operations, color processing operations;

The third extended picture set is used as an enhanced picture set.

The specific implementation of the computer-readable storage medium of the present application is basically the same as the above-mentioned embodiments of the data enhancement device and method, and will not be repeated here.

It should be noted that the serial numbers of the above-mentioned embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments. And the terms "include", "include" or any other variants thereof in this article 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, but also includes The other elements listed may also include 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 embodiments, those skilled in the art can clearly understand that the method of the above embodiments 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 disk, optical disk), 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 method described in each embodiment of the present application.

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

Claims

A data enhancement method, characterized in that the method includes:

Get the original picture;

Perform a first processing operation on the original picture to obtain a first processed picture, and merge the first processed picture with the original picture to obtain a first extended picture set, wherein the first processing operation includes at least one of the following: blur processing, Noise treatment

Perform a second processing operation on the first extended picture set to obtain the processed first extended picture set, and obtain the second extended picture set by combining the processed first extended picture set and the first extended picture set, wherein the second processing Operations include at least one of the following: moving operations, rotating operations, and filling operations;

The third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain the third extended picture set, where the third processing The operations include at least one of the following: contrast processing operations, color processing operations;

The third extended picture set is used as an enhanced picture set.
The data enhancement method according to claim 1, wherein the first processing operation is performed on the original picture to obtain a first processed picture, and the first processed picture is combined with the original picture to obtain a first extended picture set include:

Using different window sizes to perform mean blurring on the original pictures to obtain pictures corresponding to each window size, and merge the pictures corresponding to each window size with the original pictures to obtain the first merged picture set;

Perform Gaussian blur processing on each picture in the first merged picture set to obtain the Gaussian blurred picture corresponding to each picture in the first merged picture set, and combine the Gaussian blurred picture corresponding to each picture in the first merged picture set with the first merged picture Assemble and obtain a second combined picture set;

Perform Gaussian noise processing on each picture in the second merged picture set to obtain Gaussian noise processed pictures corresponding to each picture in the second merged picture set, and combine the Gaussian noise processed pictures corresponding to each picture in the second merged picture set with the second merged picture Collect and get the first extended picture set.
The data enhancement method of claim 1, wherein the second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture set is combined with The first extended picture set and the second extended picture set obtained include:

Perform mobile processing on each picture in the first extended picture set to obtain a mobile picture corresponding to each picture in the first extended picture set, and perform a mobile picture corresponding to each picture in the first extended picture set with each picture in the first extended picture set Merge to obtain the third merged picture set;

Rotate each picture in the third merged picture set by a preset angle to obtain the rotated picture corresponding to each picture in the third merged picture set, and merge the rotated picture corresponding to each picture in the third merged picture set with the third merged picture set, Obtain the fourth combined picture collection;

Filling processing is performed on the fourth combined picture set to obtain the second extended picture set.
The data enhancement method according to claim 2, wherein the second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture set is combined with The first extended picture set and the second extended picture set obtained include:

Perform mobile processing on each picture in the first extended picture set to obtain a mobile picture corresponding to each picture in the first extended picture set, and perform a mobile picture corresponding to each picture in the first extended picture set with each picture in the first extended picture set Merge to obtain the third merged picture set;

Rotate each picture in the third merged picture set by a preset angle to obtain the rotated picture corresponding to each picture in the third merged picture set, and merge the rotated picture corresponding to each picture in the third merged picture set with the third merged picture set, Obtain the fourth combined picture collection;

Filling processing is performed on the fourth combined picture set to obtain the second extended picture set.
The data enhancement method of claim 3, wherein the second processing operation further comprises a zooming operation and a trimming operation, and the second processing operation is performed on the first extended picture set to obtain the processed first extended picture The second extended picture set by combining the processed first extended picture set and the first extended picture set to obtain the second extended picture set also includes:

Perform a random zoom operation on each picture in the fourth merged picture set to obtain the zoomed picture corresponding to each picture in the fourth merged picture set, merge the zoomed picture corresponding to each picture in the fourth merged picture set with the fourth merged picture set, Obtain the fifth combined picture collection;

Perform random cropping operations on each picture in the fifth merged picture set to obtain the cropped pictures corresponding to each picture in the fifth merged picture set, and merge the cropped pictures corresponding to each picture in the fifth merged picture set with the fifth merged picture set, Get the second extended picture set.
The data enhancement method of claim 4, wherein the second processing operation further comprises a zooming operation and a trimming operation, and the second processing operation is performed on the first extended picture set to obtain the processed first extended picture Set, and combining the processed first extended picture set and the first extended picture set to obtain the second extended picture set also includes:

Perform a random zoom operation on each picture in the fourth merged picture set to obtain the zoomed picture corresponding to each picture in the fourth merged picture set, merge the zoomed picture corresponding to each picture in the fourth merged picture set with the fourth merged picture set, Obtain the fifth combined picture collection;

Perform random cropping operations on each picture in the fifth merged picture set to obtain the cropped pictures corresponding to each picture in the fifth merged picture set, and merge the cropped pictures corresponding to each picture in the fifth merged picture set with the fifth merged picture set, Get the second extended picture set.
The data enhancement method of claim 1, wherein the third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set is combined with The second extended picture set and the third extended picture set obtained include:

Perform a contrast transformation operation on each picture in the second extended picture set to obtain a contrast-transformed picture corresponding to each picture in the second extended picture set, and convert the contrast-transformed picture corresponding to each picture in the second extended picture set to each The pictures are merged to obtain the sixth merged picture set;

Perform a color transformation operation on each picture in the sixth merged picture set to obtain the color-transformed picture corresponding to each picture in the sixth merged picture set. The color-transformed picture corresponding to each picture in the sixth merged picture set is compared with the sixth picture. The combined picture sets are merged to obtain the third extended picture set.
A data enhancement device, characterized in that the device includes a memory and a processor, the memory stores a data enhancement program that can run on the processor, and when the data enhancement program is executed by the processor Implement the following steps:

Get the original picture;

Perform a first processing operation on the original picture to obtain a first processed picture, and merge the first processed picture with the original picture to obtain a first extended picture set, wherein the first processing operation includes at least one of the following: blur processing, Noise treatment

Perform a second processing operation on the first extended picture set to obtain the processed first extended picture set, and obtain the second extended picture set by combining the processed first extended picture set and the first extended picture set, wherein the second processing Operations include at least one of the following: moving operations, rotating operations, and filling operations;

The third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain the third extended picture set, where the third processing The operations include at least one of the following: contrast processing operations, color processing operations;

The third extended picture set is used as an enhanced picture set.
The data enhancement device according to claim 8, wherein the first processing operation is performed on the original picture to obtain a first processed picture, and the first processed picture is combined with the original picture to obtain a first extended picture set include:

Using different window sizes to perform mean blurring on the original pictures to obtain pictures corresponding to each window size, and merge the pictures corresponding to each window size with the original pictures to obtain the first merged picture set;

Perform Gaussian blur processing on each picture in the first merged picture set to obtain the Gaussian blurred picture corresponding to each picture in the first merged picture set, and combine the Gaussian blurred picture corresponding to each picture in the first merged picture set with the first merged picture Assemble and obtain a second combined picture set;

Perform Gaussian noise processing on each picture in the second merged picture set to obtain Gaussian noise processed pictures corresponding to each picture in the second merged picture set, and combine the Gaussian noise processed pictures corresponding to each picture in the second merged picture set with the second merged picture Collect and get the first extended picture set.
The data enhancement device of claim 8, wherein the second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture set is combined with The first extended picture set and the second extended picture set obtained include:

Perform mobile processing on each picture in the first extended picture set to obtain a mobile picture corresponding to each picture in the first extended picture set, and perform a mobile picture corresponding to each picture in the first extended picture set with each picture in the first extended picture set Merge to obtain the third merged picture set;

Rotate each picture in the third merged picture set by a preset angle to obtain the rotated picture corresponding to each picture in the third merged picture set, and merge the rotated picture corresponding to each picture in the third merged picture set with the third merged picture set, Obtain the fourth combined picture collection;

Filling processing is performed on the fourth combined picture set to obtain the second extended picture set.
The data enhancement device of claim 9, wherein the second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture set is combined with The first extended picture set and the second extended picture set obtained include:

Perform mobile processing on each picture in the first extended picture set to obtain a mobile picture corresponding to each picture in the first extended picture set, and perform a mobile picture corresponding to each picture in the first extended picture set with each picture in the first extended picture set Merge to obtain the third merged picture set;

Rotate each picture in the third merged picture set by a preset angle to obtain the rotated picture corresponding to each picture in the third merged picture set, and merge the rotated picture corresponding to each picture in the third merged picture set with the third merged picture set, Obtain the fourth combined picture collection;

Filling processing is performed on the fourth combined picture set to obtain the second extended picture set.
The data enhancement device of claim 10, wherein the second processing operation further comprises a zooming operation and a cropping operation, and the second processing operation is performed on the first extended picture set to obtain the processed first extended picture The second extended picture set by combining the processed first extended picture set and the first extended picture set to obtain the second extended picture set also includes:

Perform a random zoom operation on each picture in the fourth merged picture set to obtain the zoomed picture corresponding to each picture in the fourth merged picture set, merge the zoomed picture corresponding to each picture in the fourth merged picture set with the fourth merged picture set, Obtain the fifth combined picture collection;

Perform random cropping operations on each picture in the fifth merged picture set to obtain the cropped pictures corresponding to each picture in the fifth merged picture set, and merge the cropped pictures corresponding to each picture in the fifth merged picture set with the fifth merged picture set, Get the second extended picture set.
The data enhancement device of claim 11, wherein the second processing operation further comprises a zooming operation and a cropping operation, and the second processing operation is performed on the first extended picture set to obtain the processed first extended picture The second extended picture set by combining the processed first extended picture set and the first extended picture set to obtain the second extended picture set also includes:

Perform a random zoom operation on each picture in the fourth merged picture set to obtain the zoomed picture corresponding to each picture in the fourth merged picture set, merge the zoomed picture corresponding to each picture in the fourth merged picture set with the fourth merged picture set, Obtain the fifth combined picture collection;

Perform random cropping operations on each picture in the fifth merged picture set to obtain the cropped pictures corresponding to each picture in the fifth merged picture set, and merge the cropped pictures corresponding to each picture in the fifth merged picture set with the fifth merged picture set, Get the second extended picture set.
The data enhancement device of claim 8, wherein the third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set is combined with The second extended picture set and the third extended picture set obtained include:

Perform a contrast transformation operation on each picture in the second extended picture set to obtain a contrast-transformed picture corresponding to each picture in the second extended picture set, and convert the contrast-transformed picture corresponding to each picture in the second extended picture set to each The pictures are merged to obtain the sixth merged picture set;

Perform a color transformation operation on each picture in the sixth merged picture set to obtain the color-transformed picture corresponding to each picture in the sixth merged picture set. The color-transformed picture corresponding to each picture in the sixth merged picture set is compared with the sixth picture. The combined picture sets are merged to obtain the third extended picture set.
A computer-readable storage medium, characterized in that a data enhancement program is stored on the computer-readable storage medium, and the data enhancement program can be executed by one or more processors to implement the following steps:

Get the original picture;

Perform a first processing operation on the original picture to obtain a first processed picture, and merge the first processed picture with the original picture to obtain a first extended picture set, wherein the first processing operation includes at least one of the following: blur processing, Noise treatment

Perform a second processing operation on the first extended picture set to obtain the processed first extended picture set, and obtain the second extended picture set by combining the processed first extended picture set and the first extended picture set, wherein the second processing Operations include at least one of the following: moving operations, rotating operations, and filling operations;

The third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture set and the second extended picture set are combined to obtain the third extended picture set, where the third processing The operations include at least one of the following: contrast processing operations, color processing operations;

The third extended picture set is used as an enhanced picture set.
The computer-readable storage medium according to claim 15, wherein the first processing operation is performed on the original picture to obtain a first processed picture, and the first processed picture is combined with the original picture to obtain a first extension Picture collection includes:

Using different window sizes to perform mean blurring on the original pictures to obtain pictures corresponding to each window size, and merge the pictures corresponding to each window size with the original pictures to obtain the first merged picture set;

Perform Gaussian blur processing on each picture in the first merged picture set to obtain the Gaussian blurred picture corresponding to each picture in the first merged picture set, and combine the Gaussian blurred picture corresponding to each picture in the first merged picture set with the first merged picture Assemble and obtain a second combined picture set;

Perform Gaussian noise processing on each picture in the second merged picture set to obtain Gaussian noise processed pictures corresponding to each picture in the second merged picture set, and combine the Gaussian noise processed pictures corresponding to each picture in the second merged picture set with the second merged picture Collect and get the first extended picture set.
15. The computer-readable storage medium of claim 15, wherein the second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture The set and the first extended picture set to obtain the second extended picture set include:

Perform mobile processing on each picture in the first extended picture set to obtain a mobile picture corresponding to each picture in the first extended picture set, and perform a mobile picture corresponding to each picture in the first extended picture set with each picture in the first extended picture set Merge to obtain the third merged picture set;

Rotate each picture in the third merged picture set by a preset angle to obtain the rotated picture corresponding to each picture in the third merged picture set, and merge the rotated picture corresponding to each picture in the third merged picture set with the third merged picture set, Obtain the fourth combined picture collection;

Filling processing is performed on the fourth combined picture set to obtain the second extended picture set.
16. The computer-readable storage medium of claim 16, wherein the second processing operation is performed on the first extended picture set to obtain the processed first extended picture set, and the processed first extended picture The set and the first extended picture set to obtain the second extended picture set include:

Perform mobile processing on each picture in the first extended picture set to obtain a mobile picture corresponding to each picture in the first extended picture set, and perform a mobile picture corresponding to each picture in the first extended picture set with each picture in the first extended picture set Merge to obtain the third merged picture set;

Rotate each picture in the third merged picture set by a preset angle to obtain the rotated picture corresponding to each picture in the third merged picture set, and merge the rotated picture corresponding to each picture in the third merged picture set with the third merged picture set, Obtain the fourth combined picture collection;

Filling processing is performed on the fourth combined picture set to obtain the second extended picture set.
The computer-readable storage medium of claim 17 or 18, wherein the second processing operation further comprises a zooming operation and a cropping operation, and the second processing operation is performed on the first extended picture set to obtain the processed The first extended picture set, and the processed first extended picture set and the first extended picture set to obtain the second extended picture set further includes:

Perform a random zoom operation on each picture in the fourth merged picture set to obtain the zoomed picture corresponding to each picture in the fourth merged picture set, merge the zoomed picture corresponding to each picture in the fourth merged picture set with the fourth merged picture set, Obtain the fifth combined picture collection;

Perform random cropping operations on each picture in the fifth merged picture set to obtain the cropped pictures corresponding to each picture in the fifth merged picture set, and merge the cropped pictures corresponding to each picture in the fifth merged picture set with the fifth merged picture set, Get the second extended picture set.
16. The computer-readable storage medium of claim 16, wherein the third processing operation is performed on the second extended picture set to obtain the processed second extended picture set, and the processed second extended picture And the second extended picture set to obtain the third extended picture set includes:

Perform a contrast transformation operation on each picture in the second extended picture set to obtain a contrast-transformed picture corresponding to each picture in the second extended picture set, and convert the contrast-transformed picture corresponding to each picture in the second extended picture set to each The pictures are merged to obtain the sixth merged picture set;

Perform a color transformation operation on each picture in the sixth merged picture set to obtain the color-transformed picture corresponding to each picture in the sixth merged picture set. The color-transformed picture corresponding to each picture in the sixth merged picture set is compared with the sixth picture. The combined picture sets are merged to obtain the third extended picture set.