WO2021139179A1

WO2021139179A1 - Image sample generation method and apparatus based on local shadow special effect

Info

Publication number: WO2021139179A1
Application number: PCT/CN2020/111946
Authority: WO
Inventors: 雷晨雨; 程晓; 张国辉; 宋晨
Original assignee: 平安科技（深圳）有限公司
Priority date: 2020-07-31
Filing date: 2020-08-28
Publication date: 2021-07-15
Also published as: CN111914934B; CN111914934A

Abstract

Provided are an image sample generation method and apparatus based on a local shadow special effect, a computer device, and a storage medium. The present invention relates to artificial intelligence technology. The method comprises: calling a contrast and brightness adjustment algorithm, and carrying out contrast and brightness adjustment on an input image to obtain a corresponding dark image and a corresponding bright image; acquiring an original image size of the input image, and according to the original image size, carrying out initialization to obtain an initial fuzzy forward dodging image, and carrying out fuzzy forward dodging on the initial fuzzy forward dodging image to obtain a current fuzzy forward dodging image (S130); carrying out Gaussian blurring on the current fuzzy forward dodging image to obtain a Gaussian-blurred image (S140); and according to the dark image and the bright image corresponding to the input image and the Gaussian-blurred image, carrying out synthesis to obtain an image sample corresponding to the input image (S150). The method achieves image sample extension based on a local shadow special effect for an input image, and reduces the acquisition difficulty for acquiring an identity card image sample, and the extended image sample is close to a sample photographed in a real scene, and the problem of small samples is solved.

Description

Image sample generation method and device based on local shadow special effect

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on July 31, 2020, the application number is 202010763065.7, and the application title is "Image sample generation method and device based on partial shadow special effects", the entire content of which is incorporated by reference In this application.

Technical field

This application relates to the field of artificial intelligence image processing technology, and in particular to a method, device, computer equipment, and storage medium for generating image samples based on partial shadow special effects.

Background technique

In the ID verification project, the inventor realized that it is very difficult to collect samples of ID cards. This is mainly because ID cards are objects related to personal privacy. Due to factors such as security and privacy, the collection is very difficult. Therefore, when doing ID card-related algorithm research, there is an extreme lack of ID card image samples, which makes it more difficult to obtain sample data for training ID card recognition models.

Summary of the invention

The embodiment of the application provides an image sample generation method, device, computer equipment, and storage medium based on partial shadow special effects, aiming to solve the extreme lack of ID card image samples in the process of training ID card recognition models in the prior art, resulting in training The problem of increasing difficulty in obtaining sample data of ID card recognition model.

In the first aspect, an embodiment of the present application provides an image sample generation method based on partial shadow special effects, which includes:

Receive and save the input picture;

Call a pre-stored contrast and brightness adjustment algorithm, adjust the contrast and brightness of the input picture, to obtain a dark picture and a bright picture corresponding to the input picture;

Obtain the original image size of the input picture, initialize an initial blurred forward uniform light image according to the original image size, and perform blur forward uniform light on the initial blurred forward uniform light image to obtain the current blurred forward uniform light Light picture

Performing Gaussian blurring on the current blurred forward uniform light image to obtain a Gaussian blurred image; and

Synthesize according to the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.

In the second aspect, an embodiment of the present application provides an image sample generating device based on partial shadow special effects, which includes:

The picture receiving unit is used to receive and save the input picture;

The contrast and brightness adjustment unit is used to call a pre-stored contrast and brightness adjustment algorithm to adjust the contrast and brightness of the input picture to obtain dark pictures and bright pictures corresponding to the input picture;

The fuzzy forward uniform light image acquisition unit is configured to acquire the original image size of the input image, initialize an initial fuzzy forward uniform light image according to the original image size, and blur the initial blurred forward uniform light image Forward homogenization, get the current fuzzy positive homogenization picture;

The Gaussian blur unit is used to perform Gaussian blurring on the current blurred forward uniform light picture to obtain a Gaussian blurred picture; and

The picture synthesis unit is configured to synthesize the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.

In a third aspect, an embodiment of the present application provides a computer device, which includes a memory, a processor, and a computer program stored on the memory and running on the processor, and the processor executes the computer The program implements the image sample generation method based on the partial shadow special effect described in the first aspect.

In a fourth aspect, an embodiment of the present application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the above-mentioned first On the one hand, the method for generating image samples based on local shadow special effects.

The embodiment of the application provides an image sample generation method, device, computer equipment, and storage medium based on partial shadow special effects, including receiving and saving an input picture; calling a pre-stored contrast brightness adjustment algorithm, and performing contrast brightness on the input picture Adjust to obtain the dark picture and the bright picture corresponding to the input picture; obtain the original image size of the input picture, initialize according to the original image size to obtain an initial blur forward uniform light picture, and compare the initial blur forward Perform a fuzzy forward homogenization picture to obtain a current blurred forward homogenization picture; perform Gaussian blurring on the current fuzzy forward homogenization picture to obtain a Gaussian blurred picture; and according to the dark picture and the corresponding dark picture of the input picture The bright picture and the Gaussian blurred picture are synthesized to obtain an image sample corresponding to the input picture. It realizes the expansion of the image samples of the input pictures based on the special effects of partial shadows, which reduces the difficulty of obtaining ID card image samples, and the expanded image samples are close to the samples taken in the real scene, which solves the problem of small samples.

Description of the drawings

In order to explain the technical solutions of the embodiments of the present application more clearly, the following will briefly introduce the drawings used in the description of the embodiments. Obviously, the drawings in the following description are some embodiments of the present application. Ordinary technicians can obtain other drawings based on these drawings without creative work.

FIG. 1 is a schematic diagram of an application scenario of an image sample generation method based on partial shadow effects provided by an embodiment of the application;

2 is a schematic flowchart of an image sample generation method based on partial shadow effects provided by an embodiment of the application;

FIG. 3 is a schematic block diagram of an image sample generating apparatus based on partial shadow special effects provided by an embodiment of the application;

Fig. 4 is a schematic block diagram of a computer device provided by an embodiment of the application.

Detailed ways

The technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of them. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of this application.

It should be understood that when used in this specification and appended claims, the terms "including" and "including" indicate the existence of the described features, wholes, steps, operations, elements and/or components, but do not exclude one or The existence or addition of multiple other features, wholes, steps, operations, elements, components, and/or collections thereof.

It should also be understood that the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in the specification of this application and the appended claims, unless the context clearly indicates other circumstances, the singular forms "a", "an" and "the" are intended to include plural forms.

It should be further understood that the term "and/or" used in the specification and appended claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations .

Please refer to FIGS. 1 and 2. FIG. 1 is a schematic diagram of an application scenario of an image sample generation method based on partial shadow special effects provided by an embodiment of this application; FIG. 2 is a schematic diagram of an image sample generation method based on partial shadow special effects provided by an embodiment of this application A schematic flow chart. The method for generating image samples based on partial shadow special effects is applied to a server, and the method is executed by application software installed in the server.

As shown in Figure 2, the method includes steps S110 to S150.

S110. Receive and save the input picture.

In this embodiment, in order to expand the ID card picture sample in the picture set that includes a small number of ID card pictures, at this time, the input picture including one or more ID card pictures may be sent from the user end to the server. During specific implementation, in order to facilitate the understanding of the technical solution, an example of generating another image sample based on a single ID card picture (that is, the aforementioned input picture) is taken as an example. Obviously, one image sample is generated based on a single ID card image based on an image sample based on the local shadow special effect, and then multiple image samples are generated based on a single ID card image and repeated multiple times based on the image sample based on the local shadow special effect. In this way, the number of ID card image samples can be effectively expanded, and the problem of the extreme lack of ID card image samples in the process of training the ID card recognition model is solved.

S120: Invoke a pre-stored contrast and brightness adjustment algorithm, adjust the contrast and brightness of the input picture, to obtain a dark picture and a bright picture corresponding to the input picture.

In this embodiment, after the server obtains an input picture, the locally stored contrast and brightness adjustment algorithm can be called at this time, the input picture is darkened to obtain a dark picture, and the input picture is brightened Process to get a bright picture. The dark picture and the light picture obtained at this time are to be superimposed with the subsequent Gaussian blurred picture, so as to generate an image sample with a better simulation input picture effect.

In an embodiment, the contrast brightness adjustment algorithm is:

y _ij =[x _ij -127.5*(1-B)]*k+127.5*(1+B);

Where x _ij represents the pixel value of the pixel in the i-th row and j-th column of the input image; y _ij represents the pixel value of the pixel in the i-th row and j-th column after the contrast brightness adjustment; B is the brightness adjustment parameter, and B The value range of is [-1,1]; k is the contrast adjustment parameter, and k=tan((45+44*c)/180*π), and the value range of c is [-1,1].

In this embodiment, the value range of the brightness adjustment parameter B is [-1,1], and its function is to adjust the brightness of the image. When B takes a value between [-1,0], the input image can be adjusted. Dark, when B takes a value between [0,1], the input picture can be brightened. The value range of the contrast adjustment parameter k is [-1,1], and its function is to adjust the contrast of the image, and the value range of arctan(k) is [1,89]. Among them, when B=0, y _ij =(x _ij -127.5)*k+127.5, then the contrast brightness adjustment algorithm only adjusts the contrast; when c=0, k=1, y _ij =x _ij +255*B At this time, the contrast and brightness adjustment algorithm only adjusts the brightness.

In an embodiment, step S120 includes:

Obtaining the pixel value of each pixel in the input picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [-1, 0] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel The first adjusted pixel value of to obtain a dark picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [0, 1] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel value The second adjusted pixel value to obtain a bright picture.

In this embodiment, when the picture is adjusted according to the contrast and brightness adjustment algorithm, two pictures with different brightness levels can be obtained. Set the value of B to be [-1,0] when obtaining dark pictures, and to [0,1] when obtaining bright pictures. And in the above process, c can take any value in the value range [-1,1]. By adjusting the contrast and brightness of the input picture through the contrast and brightness adjustment algorithm, a bright picture and a dark picture with better picture quality can be obtained for subsequent picture processing.

S130. Obtain the original image size of the input image, obtain an initial fuzzy forward uniform light image by initializing the original image size, and perform blur forward uniform light on the initial blurred forward uniform light image to obtain the current blur positive image. To the uniform light picture.

In this embodiment, the uneven brightness distribution may occur in the bright picture or the dark picture, which seriously affects the picture quality and increases the difficulty of subsequent image processing. At this time, in order to eliminate the uneven brightness distribution of the picture, the image can be homogenized. In this application, the MASK uniform light algorithm can be used to process the picture to enhance high-frequency information and suppress low-frequency information, thereby enhancing the detail contrast of the picture and suppressing abnormal brightness changes of the picture, so as to achieve the purpose of uniform light.

At this time, when the MASK uniform light algorithm is used to process the picture, instead of directly processing the bright picture or the dark picture, an initial fuzzy forward uniform light picture with the same picture size as the input picture is initially generated (it can also be It is understood that the initial blurred forward uniform light picture is exactly the same as the input picture), and then the MASK uniform light is performed on it to obtain the current blurred forward uniform light picture. The obtained current fuzzy forward homogenized image can participate in the processing of dark or bright images after further processing, thereby enhancing the detail contrast of the obtained sample image and suppressing abnormal brightness changes of the sample image.

In an embodiment, as a first specific implementation of step S130, step S130 includes:

Invoking the pre-stored flat fuzzy forward uniform light algorithm, and the initial fuzzy forward uniform light image is subjected to plane fuzzy forward uniform light through the plane fuzzy forward uniform light algorithm to obtain the current blurred forward uniform light image; where , The plane formula in the plane blur forward uniform light algorithm is Y=a1X+b1, and a1 and b1 are random values; if there are pixels in the initial blur forward uniform light picture that are located in the Y=a1X+b1 In the plane, the pixel value of the corresponding pixel is set to 255; if there is a pixel in the initial blurred forward uniform light image that is outside the plane of Y=a1X+b1, then the pixel value of the corresponding pixel is set Is 0.

In this embodiment, a current fuzzy forward uniform light image is obtained during initialization, and the current fuzzy forward uniform light image can be subjected to planar MASK using the planar MASK algorithm (that is, the planar fuzzy forward uniform light algorithm) to obtain the current Blur the picture with the forward uniform light.

By performing planar mask on the current blurred forward uniform light image, each pixel in the current blurred forward uniform light image is judged as follows:

If there are pixels in the initial blurred forward uniform light picture that are located in the plane of Y=a1X+b1, the pixel value of the corresponding pixel is set to 255; if there are pixels in the initial blurred forward uniform light picture Outside the plane of Y=a1X+b1, the pixel value of the corresponding pixel is set to 0; the plane formula in the plane blur forward homogenization algorithm is Y=a1X+b1, and a1 and b1 are random values. Through the above-mentioned planar mask processing, the high frequency information in the initial blurred forward uniform light image is enhanced, the low frequency information therein is suppressed, and the abnormal brightness change of the initial blurred forward uniform light image is suppressed.

In an embodiment, as a second specific implementation of step S130, the step S130 includes:

Invoking the pre-stored curved surface fuzzy forward uniform light algorithm, the initial fuzzy forward uniform light image is subjected to planar fuzzy forward uniform light through the curved surface fuzzy forward uniform light algorithm, to obtain the current fuzzy forward uniform light image; where , The plane formula in the curved surface fuzzy forward homogenization algorithm is a*x ² +b*y ² +c*x*y+d*x+e*y=f, a, b, c, d, e are Random value; if there are pixels in the initial blurred forward homogenized image that are located in ^{the plane range of a*x 2} +b*y ² +c*x*y+d*x+e*y=f, it will correspond The pixel value of the pixel point of is set to 255; if there is a pixel point in the initial blurry forward homogenization picture located at a*x ² +b*y ² +c*x*y+d*x+e*y=f Set the pixel value of the corresponding pixel to 0 outside the plane range.

In this embodiment, a current fuzzy forward uniform light image is obtained during initialization, and the current fuzzy forward uniform light image can also be surface-masked through the curved surface MASK algorithm (that is, the curved surface fuzzy forward uniform light algorithm), to obtain The current picture is blurred forward and even light.

By performing surface MASK on the current blurred forward uniform light image, each pixel in the current blurred forward uniform light image is judged as follows:

If there are pixels in the initial blurred forward homogenization image that are located in ^{the plane range of a*x 2} + b*y ² + c*x*y+d*x+e*y=f, then the corresponding pixel points The pixel value of is set to 255; if there is a pixel in the initial blurred forward homogenization picture that is located in ^{the plane range of a*x 2} +b*y ² +c*x*y+d*x+e*y=f Outside, set the pixel value of the corresponding pixel to 0; among them, the surface formula in the surface fuzzy forward uniform light algorithm is a*x ² +b*y ² +c*x*y+d*x+e*y =f, a, b, c, d, e are random values. Through the above-mentioned curved surface MASK processing, the high frequency information in the initial blurred forward uniform light image is also enhanced, the low frequency information therein is suppressed, and the abnormal brightness change of the initial blurred forward uniform light image is suppressed. That is, in specific implementation, any one of the planar MASK algorithm or the curved surface MASK algorithm can be selected to perform the homogenization process on the initial fuzzy forward homogenization image.

S140: Perform Gaussian blurring on the current blurred forward uniform light image to obtain a Gaussian blurred image.

In this embodiment, after the current blurred forward uniform light image is obtained, in order to reduce image noise and reduce the level of detail. The Gaussian blur process of the image is the convolution of the image and the normal distribution. Since the normal distribution is also called Gaussian distribution, this technique is called Gaussian blur.

In an embodiment, step S140 includes:

Use a two-dimensional Gaussian function

And the filter kernel size is a random number in [10,100] to perform Gaussian blurring on the current blurred forward uniform light picture to obtain a Gaussian blurred picture; where (x, y) represents the current blurred forward uniform light picture The pixel coordinates of, σ is the variance of x.

Gaussian blur is defined in two-dimensional space as

At this time, after Gaussian blurring is performed on the current blurred forward homogenized image through Gaussian blurring in the two-dimensional space, a Gaussian blurred image with reduced image noise and reduced level of detail can be obtained.

S150. Synthesize according to the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.

In this embodiment, the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture are obtained in the previous steps. At this time, the bright picture can be processed by Gaussian blurred picture, and the dark picture After Gaussian blurring, the pictures are processed and then superimposed to obtain image samples corresponding to the input pictures. Through this processing, the generated picture samples have high fidelity.

In an embodiment, step S150 includes:

Performing normalization processing on the Gaussian blurred picture to obtain a normalized picture;

Acquiring a pixel matrix corresponding to the normalized picture as a first matrix, and subtracting the first matrix from the identity matrix to obtain a second matrix;

Multiplying the pixel matrix of the bright picture corresponding to the bright picture by the first matrix to obtain a first final pixel matrix, and multiplying the pixel matrix corresponding to the dark picture by the second matrix where it is located to obtain a second final pixel matrix;

The first final pixel matrix and the second final pixel matrix are added and summed to obtain an image sample corresponding to the input picture.

In this embodiment, the purpose of normalizing the Gaussian blurred picture is to process it into a weight matrix, so as to calculate the respective weights of the bright pictures and the dark pictures when they are superimposed. In specific implementation, the first matrix is taken as the weight corresponding to the bright picture, and the second matrix is taken as the weight corresponding to the dark picture, and then the first final pixel matrix and the second final pixel matrix are added and summed to obtain the input picture Corresponding image samples, the effects of the above-mentioned pictures obtained by special effects based on partial shadows are very close to samples taken in real scenes, and can even simulate samples that are difficult to collect in real scenes.

In an embodiment, after step S150, the method further includes:

Both the input picture and the image sample are sent to the user terminal.

In this embodiment, after the image sample obtained by expanding the input picture in the server, the input picture and the image sample can be sent to the user terminal, and the user terminal can train the ID card recognition model based on the expanded image sample.

During specific implementation, after the step of sending both the input picture and the image sample to the user terminal, the method further includes:

Receive and store the ID card recognition model trained by the user terminal according to the picture training set including the input picture and the image sample.

In this embodiment, when the ID card recognition model (such as convolutional neural network) to be trained is trained in the user terminal according to the image training set including the input image and image samples, the obtained ID card recognition model is sent to the server Store it.

In an embodiment, after the step of receiving and storing the ID card recognition model trained by the user terminal according to the picture training set including the input picture and the image sample, the method further includes:

Upload the parameter set corresponding to the ID card recognition model to the blockchain network.

In this embodiment, the server can be used as a blockchain node device to upload the parameter set corresponding to the ID card identification model to the blockchain network, making full use of the non-tamperable characteristics of blockchain data to realize the ID card identification model The data of the corresponding parameter set is solidified and stored. After the server re-downloads and obtains the parameter set corresponding to the ID card recognition model from the blockchain network, it can restore the ID card recognition model again according to the parameter set corresponding to the ID card recognition model.

Wherein, the corresponding summary information is obtained based on the parameter set corresponding to the ID card recognition model. Specifically, the summary information is obtained by hashing the parameter set corresponding to the ID card recognition model, for example, obtained by using the sha256 algorithm. Uploading summary information to the blockchain can ensure its security and fairness and transparency to users. The server can download the summary information from the blockchain to verify whether the parameter set corresponding to the ID card recognition model has been tampered with. The blockchain referred to in this example is a new application mode of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain, essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information for verification. The validity of the information (anti-counterfeiting) and generation of the next block. The blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.

This method realizes the expansion of image samples based on local shadow effects for input pictures, reduces the difficulty of obtaining ID card image samples, and the expanded image samples are close to samples taken in real scenes, which solves the problem of small samples.

An embodiment of the present application also provides an image sample generating device based on a partial shadow special effect, which is used to execute any embodiment of the aforementioned image sample generating method based on a partial shadow special effect. Specifically, please refer to FIG. 3, which is a schematic block diagram of an image sample generating apparatus based on partial shadow special effects according to an embodiment of the present application. The device 100 for generating image samples based on partial shadow special effects can be configured in a server.

As shown in FIG. 3, the image sample generating device 100 based on the partial shadow special effect includes: a picture receiving unit 110, a contrast brightness adjustment unit 120, a blur forward uniform light picture acquisition unit 130, a Gaussian blur unit 140, and a picture synthesis unit 150.

The picture receiving unit 110 receives and saves the input picture.

The contrast and brightness adjustment unit 120 is configured to call a pre-stored contrast and brightness adjustment algorithm to adjust the contrast and brightness of the input picture to obtain dark pictures and bright pictures corresponding to the input picture.

In an embodiment, the contrast brightness adjustment algorithm is:

y _ij =[x _ij -127.5*(1-B)]*k+127.5*(1+B);

Where x _ij represents the pixel value of the pixel in the i-th row and j-th column of the input picture; y _ij represents the pixel value of the pixel in the i-th row and j-th column after the contrast brightness adjustment; B is the brightness adjustment parameter, and B The value range of is [-1,1]; k is the contrast adjustment parameter, and k=tan((45+44*c)/180*π), and the value range of c is [-1,1].

In an embodiment, the contrast and brightness adjustment unit 120 includes:

A pixel value obtaining unit, configured to obtain the pixel value of each pixel in the input picture;

The dark picture acquisition unit is used to call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [-1, 0], so as to perform brightness contrast on the pixel value of each pixel Adjust to obtain the first adjusted pixel value corresponding to each pixel to obtain a dark picture;

The bright picture acquisition unit is used to call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [0, 1] to adjust the brightness and contrast of the pixel value of each pixel , Obtain the second adjusted pixel value corresponding to each pixel point to obtain a bright picture.

The fuzzy forward uniform light image acquisition unit 130 is configured to acquire the original image size of the input image, initialize an initial fuzzy forward uniform light image according to the original image size, and perform processing on the initial blurred forward uniform light image Blur forward uniform light, get the current fuzzy forward uniform light picture.

In an embodiment, as a first specific implementation of the fuzzy forward uniform light image acquisition unit 130, the fuzzy forward uniform light image acquisition unit 130 is further configured to:

In an embodiment, as a second specific implementation of the fuzzy forward uniform light image acquisition unit 130, the fuzzy forward uniform light image acquisition unit 130 is further configured to:

If there are pixels in the initial blurred forward homogenization image that are located in ^{the plane range of a*x 2} + b*y ² + c*x*y+d*x+e*y=f, then the corresponding pixel points The pixel value of is set to 255; if there is a pixel in the initial blurred forward homogenization picture that is located in ^{the plane range of a*x 2} +b*y ² +c*x*y+d*x+e*y=f Outside, set the pixel value of the corresponding pixel to 0; among them, the surface formula in the surface fuzzy forward uniform light algorithm is a*x ² +b*y ² +c*x*y+d*x+e*y =f, a, b, c, d, e are random values. Through the above-mentioned curved surface MASK processing, the high frequency information in the initial blurred forward uniform light image is also enhanced, the low frequency information therein is suppressed, and the abnormal brightness change of the initial blurred forward uniform light image is suppressed. That is, during specific implementation, any one of the planar MASK algorithm or the curved surface MASK algorithm can be selected to perform the homogenization process on the initial blurred forward homogenization image.

The Gaussian blurring unit 140 is configured to perform Gaussian blurring on the current blurred forward uniform light image to obtain a Gaussian blurred image.

In an embodiment, the Gaussian blur unit 140 is also used for:

Use a two-dimensional Gaussian function

Gaussian blur is defined in two-dimensional space as

The picture synthesis unit 150 is configured to synthesize the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.

In an embodiment, the picture synthesis unit 150 includes:

A normalization unit, configured to perform normalization processing on the Gaussian blurred picture to obtain a normalized picture;

A weight matrix calculation unit, configured to obtain a pixel matrix corresponding to the normalized picture as a first matrix, and subtract the first matrix from the identity matrix to obtain a second matrix;

A picture weight processing unit, configured to multiply the bright picture pixel matrix corresponding to the bright picture by the first matrix to obtain a first final pixel matrix, and multiply the pixel matrix corresponding to the dark picture by the second matrix where it is located Obtain the second final pixel matrix;

The picture weight summation and superposition unit is configured to add and sum the first final pixel matrix and the second final pixel matrix to obtain an image sample corresponding to the input picture.

In an embodiment, the device 100 for generating image samples based on partial shadow special effects further includes:

The picture sending unit is configured to send both the input picture and the image sample to the user terminal.

During specific implementation, the device 100 for generating image samples based on partial shadow special effects further includes:

The model receiving unit is used to receive and store the ID card recognition model trained by the user terminal according to the picture training set including the input picture and the image sample.

The parameter chaining unit is used to upload the parameter set corresponding to the ID card recognition model to the blockchain network.

The device realizes the expansion of image samples of input pictures based on the special effects of partial shadows, which reduces the difficulty of obtaining ID card image samples, and the expanded image samples are close to samples taken in real scenes, which solves the problem of small samples.

The above-mentioned image sample generating device based on partial shadow special effects can be implemented in the form of a computer program, and the computer program can be run on a computer device as shown in FIG. 4.

Please refer to FIG. 4, which is a schematic block diagram of a computer device according to an embodiment of the present application. The computer device 500 is a server, and the server may be an independent server or a server cluster composed of multiple servers.

Referring to FIG. 4, the computer device 500 includes a processor 502, a memory, and a network interface 505 connected through a system bus 501, where the memory may include a non-volatile storage medium 503 and an internal memory 504.

The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. When the computer program 5032 is executed, the processor 502 can execute an image sample generation method based on partial shadow effects.

The processor 502 is used to provide calculation and control capabilities, and support the operation of the entire computer device 500.

The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can make the processor 502 execute an image sample generation method based on partial shadow effects.

The network interface 505 is used for network communication, such as providing data information transmission. Those skilled in the art can understand that the structure shown in FIG. 4 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

Wherein, the processor 502 is configured to run a computer program 5032 stored in a memory to implement the method for generating image samples based on partial shadow special effects disclosed in the embodiments of the present application.

Those skilled in the art can understand that the embodiment of the computer device shown in FIG. 4 does not constitute a limitation on the specific configuration of the computer device. In other embodiments, the computer device may include more or less components than those shown in the figure. Or some parts are combined, or different parts are arranged. For example, in some embodiments, the computer device may only include a memory and a processor. In such embodiments, the structures and functions of the memory and the processor are the same as those of the embodiment shown in FIG. 4, and will not be repeated here.

It should be understood that in this embodiment of the application, the processor 502 may be a central processing unit (Central Processing Unit, CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (Digital Signal Processors, DSPs), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor.

In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium may be non-volatile or volatile. The computer-readable storage medium stores a computer program, where the computer program is executed by a processor to realize the method for generating image samples based on partial shadow special effects disclosed in the embodiments of the present application.

Those skilled in the art can clearly understand that, for the convenience and conciseness of description, the specific working process of the above-described equipment, device, and unit can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here. A person of ordinary skill in the art may realize that the units and algorithm steps of the examples described in the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of both, in order to clearly illustrate the hardware and software Interchangeability, in the above description, the composition and steps of each example have been generally described in accordance with the function. Whether these functions are executed by hardware or software depends on the specific application and design constraint conditions of the technical solution. Professionals and technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered beyond the scope of this application.

In the several embodiments provided in this application, it should be understood that the disclosed equipment, device, and method may be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, or the units with the same function may be combined into one. Units, for example, multiple units or components can be combined or integrated into another system, or some features can be omitted or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may also be electrical, mechanical or other forms of connection.

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

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

If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of this application is essentially or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium. It includes several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), magnetic disk or optical disk and other media that can store program codes.

The above are only specific implementations of this application, but the protection scope of this application is not limited to this. Anyone familiar with the technical field can easily think of various equivalents within the technical scope disclosed in this application. Modifications or replacements, these modifications or replacements shall be covered within the protection scope of this application. Therefore, the protection scope of this application shall be subject to the protection scope of the claims.

Claims

An image sample generation method based on special effects of partial shadows, which includes:

Receive and save the input picture;

Call a pre-stored contrast and brightness adjustment algorithm, adjust the contrast and brightness of the input picture, to obtain a dark picture and a bright picture corresponding to the input picture;

Obtain the original image size of the input picture, initialize an initial blurred forward uniform light image according to the original image size, and perform blur forward uniform light on the initial blurred forward uniform light image to obtain the current blurred forward uniform light Light picture

Performing Gaussian blurring on the current blurred forward uniform light image to obtain a Gaussian blurred image; and

Synthesize according to the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.
The method for generating image samples based on partial shadow special effects according to claim 1, wherein the contrast brightness adjustment algorithm is:

y ij =[x ij -127.5*(1-B)]*k+127.5*(1+B);

Where x ij represents the pixel value of the pixel in the i-th row and j-th column of the input image; y ij represents the pixel value of the pixel in the i-th row and j-th column after the contrast brightness adjustment; B is the brightness adjustment parameter, and The value range of is [-1,1]; k is the contrast adjustment parameter, and k=tan((45+44*c)/180*π), the value range of c is [-1,1];

The adjusting the contrast and brightness of the input picture to obtain a dark picture and a bright picture corresponding to the input picture includes;

Obtaining the pixel value of each pixel in the input picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [-1, 0] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel The first adjusted pixel value of to obtain a dark picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [0, 1] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel value The second adjusted pixel value to obtain a bright picture.
The method for generating image samples based on local shadow special effects according to claim 1, wherein said performing blur forward homogenization on said initial blurred forward homogenization picture to obtain a current blurred forward homogenization picture comprises:

Invoking the pre-stored flat fuzzy forward uniform light algorithm, and the initial fuzzy forward uniform light image is subjected to plane fuzzy forward uniform light through the plane fuzzy forward uniform light algorithm to obtain the current blurred forward uniform light image; where , The plane formula in the plane blur forward uniform light algorithm is Y=a1X+b1, and a1 and b1 are random values; if there are pixels in the initial blur forward uniform light picture that are located in the Y=a1X+b1 In the plane, the pixel value of the corresponding pixel is set to 255; if there is a pixel in the initial blurred forward uniform light image that is outside the plane of Y=a1X+b1, then the pixel value of the corresponding pixel is set Is 0.
The method for generating image samples based on local shadow special effects according to claim 1, wherein said performing blur forward homogenization on said initial blurred forward homogenization picture to obtain a current blurred forward homogenization picture comprises:

Invoking the pre-stored curved surface fuzzy forward uniform light algorithm, the initial fuzzy forward uniform light image is subjected to planar fuzzy forward uniform light through the curved surface fuzzy forward uniform light algorithm, to obtain the current fuzzy forward uniform light image; where , The plane formula in the curved surface fuzzy forward homogenization algorithm is a*x 2 +b*y 2 +c*x*y+d*x+e*y=f, a, b, c, d, e are Random value; if there are pixels in the initial blurred forward homogenized image that are located in the plane range of a*x 2 +b*y 2 +c*x*y+d*x+e*y=f, it will correspond The pixel value of the pixel point of is set to 255; if there is a pixel point in the initial blurry forward homogenization picture located at a*x 2 +b*y 2 +c*x*y+d*x+e*y=f Set the pixel value of the corresponding pixel to 0 outside the plane range.
The method for generating image samples based on local shadow special effects according to claim 1, wherein said performing Gaussian blurring on the current blurred forward uniform light picture to obtain a Gaussian blurred picture comprises:

Use a two-dimensional Gaussian function
And the filter kernel size is a random number in [10,100] to perform Gaussian blurring on the current blurred forward uniform light picture to obtain a Gaussian blurred picture; where (x, y) represents the current blurred forward uniform light picture The pixel coordinates of, σ is the variance of x.
The method for generating image samples based on local shadow special effects according to claim 5, wherein the synthesis is performed according to the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain a picture corresponding to the input picture Image samples of including:

Performing normalization processing on the Gaussian blurred picture to obtain a normalized picture;

Acquiring a pixel matrix corresponding to the normalized picture as a first matrix, and subtracting the first matrix from the identity matrix to obtain a second matrix;

Multiplying the pixel matrix of the bright picture corresponding to the bright picture by the first matrix to obtain a first final pixel matrix, and multiplying the pixel matrix corresponding to the dark picture by the second matrix where it is located to obtain a second final pixel matrix;

The first final pixel matrix and the second final pixel matrix are added and summed to obtain an image sample corresponding to the input picture.
The method for generating image samples based on partial shadow special effects according to claim 1, further comprising:

Both the input picture and the image sample are sent to the user terminal.
The method for generating image samples based on partial shadow special effects according to claim 1, further comprising:

Receive and store the ID card recognition model trained by the user terminal according to the picture training set including the input picture and the image sample;

Upload the parameter set corresponding to the ID card recognition model to the blockchain network.
An image sample generating device based on local shadow special effects, which includes:

The picture receiving unit is used to receive and save the input picture;

The contrast and brightness adjustment unit is used to call a pre-stored contrast and brightness adjustment algorithm to adjust the contrast and brightness of the input picture to obtain dark pictures and bright pictures corresponding to the input picture;

The fuzzy forward uniform light image acquisition unit is configured to acquire the original image size of the input image, initialize an initial fuzzy forward uniform light image according to the original image size, and blur the initial blurred forward uniform light image Forward homogenization, get the current fuzzy positive homogenization picture;

The Gaussian blur unit is used to perform Gaussian blurring on the current blurred forward uniform light picture to obtain a Gaussian blurred picture; and

The picture synthesis unit is configured to synthesize the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.
A computer device includes a memory, a processor, and a computer program stored on the memory and capable of running on the processor, wherein the processor implements the following steps when the processor executes the computer program:

Receive and save the input picture;

Call a pre-stored contrast and brightness adjustment algorithm, adjust the contrast and brightness of the input picture, to obtain a dark picture and a bright picture corresponding to the input picture;

Obtain the original image size of the input picture, initialize an initial blurred forward uniform light image according to the original image size, and perform blur forward uniform light on the initial blurred forward uniform light image to obtain the current blurred forward uniform light Light picture

Performing Gaussian blurring on the current blurred forward uniform light image to obtain a Gaussian blurred image; and

Synthesize according to the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.
The computer device according to claim 10, wherein the contrast brightness adjustment algorithm is:

y ij =[x ij -127.5*(1-B)]*k+127.5*(1+B);

Where x ij represents the pixel value of the pixel in the i-th row and j-th column of the input image; y ij represents the pixel value of the pixel in the i-th row and j-th column after the contrast brightness adjustment; B is the brightness adjustment parameter, and B The value range of is [-1,1]; k is the contrast adjustment parameter, and k=tan((45+44*c)/180*π), the value range of c is [-1,1];

The adjusting the contrast and brightness of the input picture to obtain a dark picture and a bright picture corresponding to the input picture includes;

Obtaining the pixel value of each pixel in the input picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [-1, 0] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel The first adjusted pixel value of to obtain a dark picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [0, 1] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel value The second adjusted pixel value to obtain a bright picture.
11. The computer device according to claim 10, wherein said performing the blur and forward uniform light on the initial blurry forward uniform light image to obtain the current blur forward uniform light image comprises:

Invoking the pre-stored flat fuzzy forward uniform light algorithm, and the initial fuzzy forward uniform light image is subjected to plane fuzzy forward uniform light through the plane fuzzy forward uniform light algorithm to obtain the current blurred forward uniform light image; where , The plane formula in the plane blur forward uniform light algorithm is Y=a1X+b1, and a1 and b1 are random values; if there are pixels in the initial blur forward uniform light picture that are located in the Y=a1X+b1 In the plane, the pixel value of the corresponding pixel is set to 255; if there is a pixel in the initial blurred forward uniform light image that is outside the plane of Y=a1X+b1, then the pixel value of the corresponding pixel is set Is 0.
11. The computer device according to claim 10, wherein said performing the blur and forward uniform light on the initial blurry forward uniform light image to obtain the current blur forward uniform light image comprises:

Invoking the pre-stored curved surface fuzzy forward uniform light algorithm, the initial fuzzy forward uniform light image is subjected to planar fuzzy forward uniform light through the curved surface fuzzy forward uniform light algorithm, to obtain the current fuzzy forward uniform light image; where , The plane formula in the curved surface fuzzy forward homogenization algorithm is a*x 2 +b*y 2 +c*x*y+d*x+e*y=f, a, b, c, d, e are Random value; if there are pixels in the initial blurred forward homogenized image that are located in the plane range of a*x 2 +b*y 2 +c*x*y+d*x+e*y=f, it will correspond The pixel value of the pixel point of is set to 255; if there is a pixel point in the initial blurry forward homogenization picture located at a*x 2 +b*y 2 +c*x*y+d*x+e*y=f Set the pixel value of the corresponding pixel to 0 outside the plane range.
10. The computer device according to claim 10, wherein said performing Gaussian blurring on the current blurred forward uniform light image to obtain a Gaussian blurred image comprises:

Use a two-dimensional Gaussian function
And the filter kernel size is a random number in [10,100] to perform Gaussian blurring on the current blurred forward uniform light picture to obtain a Gaussian blurred picture; where (x, y) represents the current blurred forward uniform light picture The pixel coordinates of, σ is the variance of x.
14. The computer device according to claim 14, wherein said synthesizing according to the dark picture and light picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture comprises:

Performing normalization processing on the Gaussian blurred picture to obtain a normalized picture;

Acquiring a pixel matrix corresponding to the normalized picture as a first matrix, and subtracting the first matrix from the identity matrix to obtain a second matrix;

Multiplying the pixel matrix of the bright picture corresponding to the bright picture by the first matrix to obtain a first final pixel matrix, and multiplying the pixel matrix corresponding to the dark picture by the second matrix where it is located to obtain a second final pixel matrix;

The first final pixel matrix and the second final pixel matrix are added and summed to obtain an image sample corresponding to the input picture.
The computer device according to claim 10, further comprising:

Both the input picture and the image sample are sent to the user terminal.
The computer device according to claim 10, further comprising:

Receive and store the ID card recognition model trained by the user terminal according to the image training set including input images and image samples;

Upload the parameter set corresponding to the ID card recognition model to the blockchain network.
A computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to perform the following operations:

Receive and save the input picture;

Call a pre-stored contrast and brightness adjustment algorithm, adjust the contrast and brightness of the input picture, to obtain a dark picture and a bright picture corresponding to the input picture;

Obtain the original image size of the input picture, initialize an initial blurred forward uniform light image according to the original image size, and perform blur forward uniform light on the initial blurred forward uniform light image to obtain the current blurred forward uniform light Light picture

Performing Gaussian blurring on the current blurred forward uniform light image to obtain a Gaussian blurred image; and

Synthesize according to the dark picture and the bright picture corresponding to the input picture and the Gaussian blurred picture to obtain an image sample corresponding to the input picture.
18. The computer-readable storage medium of claim 18, wherein the contrast brightness adjustment algorithm is:

y ij =[x ij -127.5*(1-B)]*k+127.5*(1+B);

Where x ij represents the pixel value of the pixel in the i-th row and j-th column of the input picture; y ij represents the pixel value of the pixel in the i-th row and j-th column after the contrast brightness adjustment; B is the brightness adjustment parameter, and B The value range of is [-1,1]; k is the contrast adjustment parameter, and k=tan((45+44*c)/180*π), the value range of c is [-1,1];

The adjusting the contrast and brightness of the input picture to obtain a dark picture and a bright picture corresponding to the input picture includes;

Obtaining the pixel value of each pixel in the input picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [-1, 0] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel The first adjusted pixel value of to obtain a dark picture;

Call the contrast and brightness adjustment algorithm, and set the value range of the brightness adjustment parameter in the contrast and brightness adjustment algorithm to [0, 1] to adjust the brightness and contrast of the pixel value of each pixel to obtain the corresponding pixel value The second adjusted pixel value to obtain a bright picture.
18. The computer-readable storage medium according to claim 18, wherein said performing blur and forward dodging on the initial blurry forward dodging picture to obtain a current blurry forward dodging picture comprises:

Invoking the pre-stored flat fuzzy forward uniform light algorithm, and the initial fuzzy forward uniform light image is subjected to plane fuzzy forward uniform light through the plane fuzzy forward uniform light algorithm to obtain the current blurred forward uniform light image; where , The plane formula in the plane blur forward uniform light algorithm is Y=a1X+b1, and a1 and b1 are random values; if there are pixels in the initial blur forward uniform light picture that are located in the Y=a1X+b1 In the plane, the pixel value of the corresponding pixel is set to 255; if there is a pixel in the initial blurred forward uniform light image that is outside the plane of Y=a1X+b1, then the pixel value of the corresponding pixel is set Is 0.