CN115705636A - Privacy protection method, device, equipment and medium - Google Patents

Abstract

According to the embodiment of the application, before the RGBA image generated based on the original RGB image containing the user and the mask image corresponding to the user is uploaded to a cloud server, the background area in the original RGB image is replaced by the preset background, so that the background privacy of the user is protected, and the risk of privacy disclosure is reduced.

Description

A privacy protection method, device, equipment and medium

technical field

The present application relates to the technical field of image processing, and in particular to a privacy protection method, device, equipment and medium.

Background technique

With the rapid development of virtual reality technology, remote multi-person virtual meetings, virtual group photos, and virtual stages have enriched people's lives, and in the process of remote multi-person virtual meetings, virtual group photos, and virtual stages, it is necessary to synthesize multiple in the same background.

In the prior art, in the process of synthesizing multiple people in the same background, it is necessary to obtain the user's original RGB image and the user's corresponding mask image (MASK) on the device side, and based on the user's RGB image and mask image The RGBA image is obtained, and the RGBA image is uploaded to the cloud server, so that the cloud server synthesizes the images of the multiple users in the same background based on the RGBA images of each user. Since the original RGB image can be parsed from the RGBA image through image decomposition, and the original RGB image carries the real environment information of the user, that is to say, the original RGB image contains real background information, which cannot guarantee the privacy of the user's background. Risk of privacy breach.

Contents of the invention

The present application provides a privacy protection method, device, device and medium, which are used to solve the problem that the RGBA generated directly based on the original RGB image including the background area and the corresponding mask image in the prior art cannot guarantee the privacy of the user's background, and there is privacy. Dangerous problem of leakage.

The present application provides a privacy protection method, the method comprising:

Obtain the original RGB image containing the user, and obtain the mask image corresponding to the user based on the portrait segmentation algorithm;

According to the original RGB image, the mask image and the target background image, determine the target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image;

Obtain a target RGBA image based on the target RGB image and the mask image.

The present application provides a privacy protection device, the device comprising:

The obtaining module is used to obtain the original RGB image containing the user, and based on the portrait segmentation algorithm, obtain the mask image corresponding to the user;

A replacement module, configured to determine, according to the original RGB image, the mask image and the target background image, a target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image;

A processing module, configured to obtain a target RGBA image based on the target RGB image and the mask image.

The present application also provides an electronic device, the electronic device includes at least a processor and a memory, and the processor is configured to implement the steps of any one of the privacy protection methods described above when executing a computer program stored in the memory.

The present application also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the steps of any one of the above-mentioned privacy protection methods are implemented.

In the embodiment of the present application, the original RGB image containing the user is obtained, and based on the portrait segmentation algorithm, the mask image corresponding to the user is obtained, and the original RGB image is determined according to the original RGB image, the mask image and the target background image. The background area in is replaced with the target RGB image obtained from the area corresponding to the target background image, and the target RGBA image is obtained based on the target RGB image and the mask image. Because in the embodiment of the present application, before uploading the RGBA image generated based on the original RGB image containing the user and the mask image corresponding to the user to the cloud server, first replace the background area in the original RGB image with the preset background, and then The user background privacy is protected and the risk of privacy leakage is reduced.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings that need to be used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For Those of ordinary skill in the art can also obtain other drawings based on these drawings without any creative effort.

FIG. 1 is a schematic diagram of the process of a privacy protection method provided by some embodiments of the present application;

Fig. 2 a is the display schematic diagram of synthetic RGBA image in the related art;

Fig. 2b is a display schematic diagram based on RGBA decomposition in the related art;

Figure 3a is a schematic display of a synthetic RGBA image that is beneficial to privacy protection provided by some embodiments of the present application;

Figure 3b is a schematic display of a synthetic RGBA image decomposition that is beneficial to privacy protection provided by some embodiments of the present application;

Fig. 4a is a schematic diagram of a process of portrait segmentation and synthesis in the related art;

Fig. 4b is a schematic diagram of a process of portrait segmentation and synthesis provided in some embodiments of the present application;

Fig. 5 is a schematic structural diagram of a privacy protection device provided by some embodiments of the present application;

Fig. 6 is a schematic structural diagram of an electronic device provided by some embodiments of the present application.

Detailed ways

In order to make the purpose, technical solution and advantages of the application clearer, the application will be further described in detail below in conjunction with the accompanying drawings. Apparently, the described embodiments are only some of the embodiments of the application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

In the embodiment of the present application, the original RGB image containing the user is obtained, and based on the portrait segmentation algorithm, the mask image corresponding to the user is obtained, and the original RGB image is determined according to the original RGB image, the mask image and the target background image. The background area in is replaced with the target RGB image obtained from the area corresponding to the target background image, and the target RGBA image is obtained based on the target RGB image and the mask image. Because in the embodiment of the present application, before uploading the RGBA image generated based on the original RGB image containing the user and the mask image corresponding to the user to the cloud server, first replace the background area in the original RGB image with the preset background, and then The user background privacy is protected and the risk of privacy leakage is reduced.

In order to protect user background privacy and reduce the risk of privacy leakage, embodiments of the present application provide a privacy protection method, device, device, and medium.

Figure 1 is a schematic diagram of the process of a privacy protection method provided by some embodiments of the present application, the process includes the following steps:

S101: Obtain an original RGB image including a user, and obtain a mask image corresponding to the user based on a portrait segmentation algorithm.

The privacy protection method provided in the embodiment of the present application is applied to an electronic device, and the electronic device may be a device such as a smart terminal, a PC, or a server.

When there is a need to synthesize multiple images on the same background, such as multi-person virtual meetings, virtual group photos, virtual stages, etc., the electronic device needs to obtain the original RGB image containing the user for each user who is synthesizing, Wherein, the user is any user who needs to be synthesized in the same background, and the original RGB image may contain one user or may contain multiple users. After obtaining the original RGB image containing the user, in this application, based on the original RGB image and a portrait segmentation algorithm, a mask image corresponding to the user is obtained.

Specifically, when the mask image corresponding to the user is obtained through the portrait segmentation algorithm, it can be realized based on an encoder-decoder model, which includes a detail feature extraction module, a global feature extraction module, a feature fusion module and a pixel classification module. In the process of obtaining the mask image corresponding to the user, the original RGB image is first input into the model, and the detail feature extraction module performs two-dimensional convolution (Conv2D) and depth separable convolution (DSConv) on the original RGB image. ) processing to obtain the first feature map, the global feature module performs a series of convolution, upsampling and other operations on the first feature map to obtain the second feature map, the feature fusion module uses the second feature map and the first feature map The feature fusion is performed by convolution to obtain the third feature map, and finally the pixel classification module performs convolution processing on the third feature map, and performs pixel classification by softmax to obtain the mask image corresponding to the user.

S102: According to the original RGB image, the mask image and the target background image, determine a target RGB image obtained by replacing a background area in the original RGB image with an area corresponding to the target background image.

Since the original RGB image includes not only the area corresponding to the user, but also the background area in the original RGB image, wherein the background area includes the real background environment information of the user when the original RGB image is collected, so for To protect the user's background privacy, in this application, the background area contained in the original RGB image can be replaced.

In order to replace the background area, in this application, the target background image can be pre-saved in the electronic device, or the electronic device can obtain the target background image after obtaining the original RGB image containing the user and the corresponding mask image of the user , the target background image contains the information of the background to be replaced. Specifically, according to the original RGB image, the mask image and the target background image, replace the background area contained in the original RGB image, and determine to replace the background area in the original RGB image with the target background image The corresponding region gets the target RGB image. Wherein, the target RGB image includes an area corresponding to the user and a replaced background area, and the replaced background area includes background information in the target background image.

S103: Obtain a target RGBA image based on the target RGB image and the mask image.

In order to realize the synthesis of multi-person images in the same background, that is to say, in order to ensure the realization of technologies such as remote multi-person virtual meeting, virtual group photo, virtual stage, etc., after the electronic device obtains the target RGB image, it can And the mask image, obtain the target RGBA image, wherein, the target RGBA image is a three-channel image, respectively R channel, G channel and B channel, the mask image is a one-channel image, and the target RGBA image is a four-channel image .

In addition, after the target RGBA image is obtained, the target RGBA image is uploaded to the cloud server. After the electronic device uploads the target RGBA image to the cloud server, the cloud server obtains the target RGBA image corresponding to each user, decomposes the target RGBA image into a target RGB image and a mask image, and based on the scene synthesis algorithm and each user's target The RGB image is combined with the images of each user in the same background to obtain a combined image, and the combined image is displayed on the display device.

Wherein, the process of synthesizing each user in the same background based on the scene synthesis algorithm and each user's target RGB image is a prior art, and will not be repeated here.

Due to different application scenarios, for each user, the required number of original RGB images containing the user is also different, and therefore, the number of mask images corresponding to the user is also different. For example, if the application scenario is a virtual group photo, you only need to obtain one original RGB image for each user. Therefore, the number of mask images corresponding to each user is also one, and the number of final target RGBA images is also one. If the application scenario is a virtual stage, etc., for each user, it is necessary to obtain multiple original RGB images containing the user, obtain multiple mask images corresponding to the user, and finally obtain multiple target RGBA images. In the case where multiple original RGB images containing the user are required in the application scenario such as a virtual stage, the principle of background area replacement for each original RGB image, and the determination of the corresponding target RGBA based on each original RGB image and the corresponding mask image The principles of the images are the same. Therefore, for the convenience of description, in the embodiment of the present application, the application scenario of any obtained original RGB image will be described.

In order to facilitate uploading to the cloud server, in this application, if the application scenario is a virtual stage, etc., after obtaining multiple target RGBA images of each user, the multiple target RGBA images are encoded to generate a video stream, and the video stream Upload to the cloud server, wherein the technology of encoding multiple target RGBA images to generate a video stream is an existing technology, and will not be repeated here.

Because in the embodiment of the present application, before uploading the RGBA image generated based on the original RGB image containing the user and the mask image corresponding to the user to the cloud server, first replace the background area in the original RGB image with the preset background, and then The user background privacy is protected and the risk of privacy leakage is reduced.

In order to replace the background area in the original RGB image, and then protect the background privacy of the user, on the basis of the above-mentioned embodiments, in the embodiment of the present application, the original RGB image, the mask image and the target background image , it is determined that the background region in the original RGB image is replaced with the region corresponding to the target background image, and the target RGB image obtained includes:

For each pixel point, according to I=α*f+(1-α)*b, determine the target pixel value corresponding to the pixel point, wherein, the α is the pixel value corresponding to the pixel point in the mask image, f is the pixel value of the pixel in the original RGB image, and b is the pixel value corresponding to the pixel in the target background image.

According to the target pixel value corresponding to each pixel point, the corresponding target RGB image is determined.

In order to replace the background area in the original RGB image, in this application, for each pixel contained in the original RGB image, replace the pixel value corresponding to each pixel in the background area with the corresponding pixel in the target image The pixel value corresponding to the point. Specifically, for each pixel, according to I=α*f+(1-α)*b, determine the target pixel value corresponding to the pixel, where the α is the pixel value corresponding to the pixel in the mask image , f is the pixel value of the pixel in the original RGB image, and b is the pixel value corresponding to the pixel in the target background image. Wherein, the pixel value of the pixel point corresponding to the portrait in the mask image is 255, and the pixel value of the pixel point corresponding to the non-portrait is 0.

For the convenience of description, if the original RGB image contains four pixels, namely pixel 1, pixel 2, pixel 3 and pixel 4, the pixel 1, pixel 2, pixel 3 and pixel 4 are in The corresponding pixel values in the original RGB image are 0, 255, 0, 255 respectively, and the corresponding pixel values of pixel 1, pixel 2, pixel 3 and pixel 4 in the mask image are 255, 255, respectively, 0, 0, the corresponding pixel values of pixel 1, pixel 2, pixel 3 and pixel 4 in the target background image can be 255, 0, 255, 255, then pixel 1, pixel 2, pixel The corresponding target pixel values of 3 and 4 in the target background image are 0, 255, 255, 255 respectively.

In order to obtain the target RGBA image, on the basis of the above-mentioned embodiments, in the embodiment of the present application, the obtaining the target RGBA image based on the target RGB image and the mask image includes:

Obtain the target pixel value corresponding to each pixel according to the pixel value of the pixel contained in the target RGB image in each channel, the pixel value of the pixel contained in the mask image and a preset function;

A target RGBA image is determined according to the target pixel value corresponding to each pixel point.

In this application, in the process of obtaining the target RGBA image according to the target RGB image and the mask image, it can be based on the pixel values of the pixels contained in the target RGB image in the R channel, the pixels contained in the target RGB image in The pixel value of the G channel, the pixel value of the pixel contained in the target RGB image in the B channel, the pixel value of the pixel contained in the mask image, and the preset function to obtain the target pixel value corresponding to each pixel, and then The target RGBA image is determined according to the target pixel value corresponding to each pixel, that is, for each pixel, the target pixel value corresponding to the pixel is the pixel value corresponding to the pixel in the target RGBA image.

In order to obtain the target pixel value corresponding to each pixel, on the basis of the above-mentioned embodiments, in the embodiment of the present application, according to the pixel value of the pixel contained in the target RGB image in each channel, the mask The pixel values of the pixels contained in the image and the preset functions to obtain the target pixel values corresponding to each pixel include:

For each pixel, according to the pixel value of each channel of the pixel in the target RGB image, the pixel value of the pixel in the mask image, the weight value corresponding to each channel, and the corresponding weight value of the mask image The weight value is to determine the weight sum corresponding to the pixel point, and determine the weight sum as the target pixel value corresponding to the pixel point.

Since the RGB image has three channels of R, G, and B, each pixel of the RGB image has a corresponding pixel value in these three channels, and the pixel value corresponding to the pixel in the RGB image is based on the pixel in R, G The pixel values corresponding to the three channels of R, B, and B are determined. Specifically, in the process of determining the pixel value corresponding to the pixel point in the RGB image based on the pixel values corresponding to the R, G, and B channels, the RGB function is generally used. Wherein, the RGB function is Red+256*Green+65536*Blue, and the pixel value corresponding to the pixel point in the RGB image is determined according to the RGB function, wherein, Red is the pixel value corresponding to the pixel point in the R channel, and Green is The pixel value corresponding to the pixel point in the G channel, and Blue is the pixel value corresponding to the pixel point in the B channel.

In this application, a four-channel target RGBA image can be generated based on a three-channel target RGB image and a one-channel mask image. Specifically, in the process of generating the target RGBA image, for each pixel, the pixel corresponds to a pixel value in the channel corresponding to the R channel, G channel, B channel, and mask image. The pixel values corresponding to the four channels, determine the target pixel value corresponding to the pixel, and then determine the target RGBA image based on the target pixel, that is, for each pixel, according to each pixel in the target RGB image The pixel value of the channel, the pixel value of the pixel in the mask image, the weight value corresponding to each channel and the weight value corresponding to the mask image, determine the weight sum of the pixel values, and determine the weight sum as the pixel The target pixel value corresponding to the point.

For example, the pixel value of the R channel of pixel A in the target RGB image is 255, the pixel value of the G channel of pixel A in the target RGB image is 0, and the pixel value of the B channel of pixel A in the target RGB image is 0, the pixel value of pixel A in the mask image is 0, and the weight values corresponding to each channel of R, G, and B are 1, 256, 65536, and 1 respectively, then the pixel value of the pixel A in the target RGBA image The value is 255.

In addition, the target RGBA image can also be determined based on the merge function of opencv, the matrix corresponding to the target RGB image, and the matrix corresponding to the mask image, wherein each element in the matrix corresponding to the target RGB image is each pixel in the target RGB image The corresponding pixel value, each element in the matrix corresponding to the mask image is the pixel value corresponding to each pixel in the mask image, wherein, based on the merge function of opencv, the matrix corresponding to the target RGB image, and the matrix corresponding to the mask image , the process of determining the target RGBA image is a prior art, and will not be repeated here.

Figure 2a is a schematic display diagram of a composite RGBA image in the related art, Figure 2b is a schematic display diagram based on RGBA decomposition in the related art, and Figure 3a is a schematic display diagram of a composite RGBA image that is beneficial to privacy protection provided by some embodiments of the present application, Figure 3b is a schematic display of a synthetic RGBA image decomposition that is beneficial to privacy protection provided by some embodiments of the present application, and is now described for Figure 2a, Figure 2b, Figure 3a and Figure 3b:

In Fig. 2a, RGB is the obtained original RGB image including the user, and α is the mask image corresponding to the user. In the prior art, the RGBA image is obtained directly based on the original RGB image and the mask image. The RGBA image in Figure 2b is decomposed into the original RGB image containing the background area, which contains the user's background privacy, and the mask image.

In Figure 3a, RGB is the obtained RGB image containing the replaced pixels in the background area of the original RGB image of the user, that is, the target RGB image, and α is the mask image corresponding to the user. In this application, based on the replaced pixels in the background area The obtained target RGB image and the mask image obtain the target RGBA image, that is to say, the target RGBA image is obtained based on the target RGB image and the mask image, and the target RGBA image in Figure 3b is decomposed into the obtained after the pixels in the background area are replaced The target RGB image and the mask image, the target RGB image obtained after replacing pixels in the background area does not contain the user's background privacy.

Fig. 4a is a schematic diagram of a process of portrait segmentation and synthesis in the related art, and Fig. 4b is a schematic diagram of a process of portrait segmentation and synthesis provided in some embodiments of the present application. Now, we will describe Fig. 4a and Fig. 4b:

Taking the application scene as an example where multiple original RGB images containing the user are needed, such as a virtual stage, in the prior art, the electronic device obtains multiple original RGB images containing the user, and uses the portrait segmentation algorithm and the original RGB image segmentation Foreground portrait, obtain multiple mask images, for each original RGB image and the mask image corresponding to the original RGB image, form the original RGB image and the mask image into a target RGBA image, and encode the target RGBA image to generate The video stream is uploaded to the cloud server. The cloud server receives the video stream corresponding to each user, and for each video stream, obtains the target RGBA image based on the video stream, decomposes the target RGBA image to generate the original RGB image and mask image, and uses the scene synthesis algorithm, the original RGB image and mask image, synthesize multiple images into the same background, and push the synthesized image to the display terminal for display.

In the embodiment of the present application, the electronic device obtains multiple original RGB images containing the user, and uses the portrait segmentation algorithm and the original RGB image to segment the foreground portrait to obtain multiple mask images. For each original RGB image, the original RGB image Perform pixel replacement in the background area to generate a target RGB image, combine the target RGB image and the mask image into a target RGBA image, and encode the target RGBA image to generate a video stream and upload it to the cloud server. The cloud server receives the video stream corresponding to each user, for each video stream, obtains the target RGBA image based on the video stream, decomposes the target RGBA image to generate the target RGB image and mask image, and uses the scene synthesis algorithm, the target RGB image and mask image, multiple people are synthesized into the same background, and the synthesized image is pushed to the display terminal for display.

Fig. 5 is a schematic structural diagram of a privacy protection device provided by some embodiments of the present application, the device includes:

The obtaining module 501 is used to obtain the original RGB image including the user, and obtain the mask image corresponding to the user based on the portrait segmentation algorithm;

A replacement module 502, configured to, according to the original RGB image, the mask image, and the target background image, determine a target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image;

The processing module 503 is configured to obtain a target RGBA image based on the target RGB image and the mask image.

In a possible implementation manner, the replacement module 502 is specifically configured to, for each pixel point, determine the target pixel value corresponding to the pixel point according to I=α*f+(1-α)*b, wherein, Said α is the corresponding pixel value of the pixel in the mask image, f is the pixel value of the pixel in the original RGB image, and b is the corresponding pixel value of the pixel in the target background image; according to each pixel The corresponding target pixel value is used to determine the corresponding target RGB image.

In a possible implementation manner, the processing module 503 is specifically configured to, according to the pixel values of the pixels contained in the target RGB image in each channel, the pixel values of the pixels contained in the mask image, and the preset The function is set to obtain the target pixel value corresponding to each pixel point; according to the target pixel value corresponding to each pixel point, the target RGBA image is determined.

In a possible implementation manner, the processing module 503 is specifically configured to, for each pixel, according to the pixel value of each channel of the pixel in the target RGB image, the pixel in the mask image The pixel value, the weight value corresponding to each channel and the weight value corresponding to the mask image, determine the weight sum corresponding to the pixel point, and determine the weight sum as the target pixel value corresponding to the pixel point.

On the basis of the foregoing embodiments, some embodiments of the present application also provide an electronic device, as shown in FIG. 6 , including: a processor 601, a communication interface 602, a memory 603, and a communication bus 604, wherein , the communication interface 602 and the memory 603 communicate with each other through the communication bus 604 .

A computer program is stored in the memory 603, and when the program is executed by the processor 601, the processor 601 is made to perform the following steps:

Obtain the original RGB image containing the user, and obtain the mask image corresponding to the user based on the portrait segmentation algorithm;

According to the original RGB image, the mask image and the target background image, determine the target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image;

Obtain a target RGBA image based on the target RGB image and the mask image.

Further, the processor 601 is also configured for each pixel according to

I=α*f+(1-α)*b, determine the target pixel value corresponding to the pixel point, wherein, the α is the pixel value corresponding to the pixel point in the mask image, and f is the pixel value corresponding to the pixel point in the original RGB image The pixel value of the pixel point, b is the pixel value corresponding to the pixel point in the target background image; according to the target pixel value corresponding to each pixel point, the corresponding target RGB image is determined.

Further, the processor 601 is further configured to obtain the pixel values of the pixels contained in the target RGB image in each channel, the pixel values of the pixels contained in the mask image, and a preset function to obtain each The target pixel value corresponding to each pixel point; according to the target pixel value corresponding to each pixel point, the target RGBA image is determined.

Further, the processor 601 is further configured to, for each pixel, according to the pixel value of the pixel in the mask image, each channel The corresponding weight value and the weight value corresponding to the mask image determine the weight sum corresponding to the pixel point, and determine the weight sum as the target pixel value corresponding to the pixel point.

The communication bus mentioned in the above server may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (Extended Industry Standard Architecture, EISA) bus or the like. The communication bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

The communication interface 602 is used for communication between the electronic device and other devices.

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

Above-mentioned processor can be general-purpose processor, comprises central processing unit, network processor (NetworkProcessor, NP) etc.; Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

On the basis of the above-mentioned embodiments, some embodiments of the present application also provide a computer-readable storage medium, where a computer program executable by an electronic device is stored in the computer-readable storage medium. When the program is in the When running on the electronic device, the following steps are implemented when the electronic device is executed:

A computer program is stored in the memory, and when the program is executed by the processor, the processor is made to perform the following steps:

Obtain the original RGB image containing the user, and obtain the mask image corresponding to the user based on the portrait segmentation algorithm;

According to the original RGB image, the mask image and the target background image, determine the target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image;

Obtain a target RGBA image based on the target RGB image and the mask image.

Further, according to the original RGB image, the mask image and the target background image, determining the target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image includes:

For each pixel point, according to I=α*f+(1-α)*b, determine the target pixel value corresponding to the pixel point, wherein, the α is the pixel value corresponding to the pixel point in the mask image, f is the pixel value of the pixel in the original RGB image, and b is the corresponding pixel value of the pixel in the target background image; according to the target pixel value corresponding to each pixel, determine the corresponding target RGB image.

Further, said obtaining the target RGBA image based on the target RGB image and the mask image includes:

Obtain the target pixel value corresponding to each pixel according to the pixel value of the pixel contained in the target RGB image in each channel, the pixel value of the pixel contained in the mask image and a preset function;

A target RGBA image is determined according to the target pixel value corresponding to each pixel point.

Further, according to the pixel values of the pixels contained in the target RGB image in each channel, the pixel values of the pixels contained in the mask image and the preset function, obtaining the target pixel value corresponding to each pixel includes :

For each pixel, according to the pixel value of each channel of the pixel in the target RGB image, the pixel value of the pixel in the mask image, the weight value corresponding to each channel, and the corresponding weight value of the mask image The weight value is to determine the weight sum corresponding to the pixel point, and determine the weight sum as the target pixel value corresponding to the pixel point.

Because in the embodiment of the present application, before uploading the RGBA image generated based on the original RGB image containing the user and the mask image corresponding to the user to the cloud server, first replace the background area in the original RGB image with the preset background, and then The user background privacy is protected and the risk of privacy leakage is reduced.

Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to the present application. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

Obviously, those skilled in the art can make various changes and modifications to the application without departing from the spirit and scope of the application. In this way, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims (10)

1. A privacy protection method, characterized in that the method comprises: Obtain the original RGB image containing the user, and obtain the mask image corresponding to the user based on the portrait segmentation algorithm; According to the original RGB image, the mask image and the target background image, determine the target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image; Obtain a target RGBA image based on the target RGB image and the mask image. 2. The method according to claim 1, wherein, according to the original RGB image, the mask image and the target background image, it is determined to replace the background area in the original RGB image with the target background image corresponding The regions of the resulting target RGB image include: For each pixel point, according to I=α*f+(1-α)*b, determine the target pixel value corresponding to the pixel point, wherein, the α is the pixel value corresponding to the pixel point in the mask image, f Be the pixel value of the pixel point in the original RGB image, b is the pixel value corresponding to the pixel point in the target background image; According to the target pixel value corresponding to each pixel point, the corresponding target RGB image is determined. 3. The method according to claim 1, wherein said obtaining a target RGBA image based on said target RGB image and said mask image comprises: Obtain the target pixel value corresponding to each pixel according to the pixel value of the pixel contained in the target RGB image in each channel, the pixel value of the pixel contained in the mask image and a preset function; A target RGBA image is determined according to the target pixel value corresponding to each pixel point. 4. The method according to claim 3, characterized in that, according to the pixel values of the pixels contained in the target RGB image in each channel, the pixel values of the pixels contained in the mask image and the preset function , obtaining the target pixel value corresponding to each pixel includes: For each pixel, according to the pixel value of each channel of the pixel in the target RGB image, the pixel value of the pixel in the mask image, the weight value corresponding to each channel, and the corresponding weight value of the mask image The weight value is to determine the weight sum corresponding to the pixel point, and determine the weight sum as the target pixel value corresponding to the pixel point. 5. A privacy protection device, characterized in that the device comprises: The obtaining module is used to obtain the original RGB image containing the user, and based on the portrait segmentation algorithm, obtain the mask image corresponding to the user; A replacement module, configured to, according to the original RGB image, the mask image and the target background image, determine a target RGB image obtained by replacing the background area in the original RGB image with the area corresponding to the target background image; A processing module, configured to obtain a target RGBA image based on the target RGB image and the mask image. 6. The device according to claim 5, wherein the replacement module is specifically configured to, for each pixel, determine the target corresponding to the pixel according to I=α*f+(1-α)*b Pixel value, wherein, said α is the pixel value corresponding to the pixel point in the mask image, f is the pixel value of the pixel point in the original RGB image, and b is the pixel value corresponding to the pixel point in the target background image; According to the target pixel value corresponding to each pixel point, the corresponding target RGB image is determined. 7. The device according to claim 5, wherein the processing module is specifically configured to use pixel values in each channel according to the pixel points contained in the target RGB image, and the pixel points contained in the mask image to The pixel value and the preset function to obtain the target pixel value corresponding to each pixel point; according to the target pixel value corresponding to each pixel point, determine the target RGBA image. 8. The device according to claim 7, wherein the processing module is specifically configured to, for each pixel, according to the pixel value of each channel of the pixel in the target RGB image, the mask image The pixel value of the pixel, the weight value corresponding to each channel and the weight value corresponding to the mask image, determine the weight sum corresponding to the pixel point, and determine the weight sum as the target pixel value corresponding to the pixel point . 9. An electronic device, characterized in that the electronic device comprises a processor, and the processor is configured to implement the steps of the method according to any one of claims 1-4 when executing the computer program stored in the memory. 10. A computer-readable storage medium, characterized in that it stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1-4 are implemented.

Images