CN117611460A - Face image fusion method, device, equipment and storage medium - Google Patents

Abstract

The invention relates to the field of images, and discloses a face image fusion method, device, equipment and storage medium. The method includes: acquiring a first face image and a second face image; determining the face skin color area of the first face image through a skin color detection algorithm, and converting the color space of the face skin color area; The area is separated from the first face image; the face skin color value of the second face image is extracted, the skin color of the separated face skin color area is adjusted, and the skin color adjusted face skin color area is compared with the first person after separation. Other areas in the face image are synthesized to obtain the third face image; based on the third face image and the second face image, the target face image is obtained through the trained generative adversarial network. In embodiments of the present invention, when faces with different skin colors are fused, skin color fusion can be made more natural and the degree of skin color fusion can be improved.

Description

Face image fusion method, device, equipment and storage medium

Technical field

The present invention relates to the field of images, and in particular to a face image fusion method, device, equipment and storage medium.

Background technique

With the development of various technologies, face processing technology in images is becoming more and more widespread. All major camera software have functions such as beautification, stickers, hairstyle changing, and fusion. Among them, the face image fusion technology used in fusion is to fuse two different face images into one face image. The newly obtained fused image Facial features of the original face image are retained.

The current technology used for face fusion is based on the method of Generative Adversarial Networks (GANs). However, when GANs fuse faces with different skin colors, the skin color fusion is unnatural. Therefore, the skin color fusion degree of existing face image fusion methods is low.

Contents of the invention

The main purpose of the present invention is to solve the technical problem of low skin color fusion when fusing facial images.

A first aspect of the present invention provides a face image fusion method. The face image fusion method includes:

Obtain a first face image and a second face image, the first face image is an unprocessed original face image, and the second face image is a reference face for changes in the first face image image;

Determine the facial skin color area of the first facial image through a skin color detection algorithm, and convert the color space of the human face skin color area;

Separate the face skin color area from the first face image according to the color space converted face skin color area and the preset color attribute range;

Extract the face skin color value of the second face image, perform skin color adjustment on the separated face skin color area according to the face skin color value, and compare the skin color adjusted face skin color area with the separated face skin color area. Synthesize other areas in the first face image to obtain a third face image;

According to the third face image and the second face image, the target face image is obtained through the trained generative adversarial network.

Optionally, in a first implementation manner of the first aspect of the present invention, the preset color attribute range includes a preset hue range and a preset saturation range;

Separating the face skin color area from the first face image according to the color space converted face skin color area and the preset color attribute range includes:

Select the target pixel in the skin color area of the human face after color space conversion according to the preset hue range and the preset saturation range;

According to the target pixels, the target face skin color area is obtained;

Separate the target face skin color area from the first face image.

Optionally, in a second implementation manner of the first aspect of the present invention, the target pixels in the human face skin color area after color space conversion are selected based on the preset hue range and the preset saturation range. include:

Generate a mask according to the preset hue range and the preset saturation range;

Perform bit operations on the mask and the first face image to obtain target pixels in the skin color area of the face after color space conversion.

Optionally, in a third implementation manner of the first aspect of the present invention, the facial skin color value of the second facial image is extracted, and the separated human face is extracted based on the facial skin color value. Perform skin color adjustment on the face skin color area, and synthesize the adjusted face skin color area with other areas in the separated first face image to obtain a third face image including:

Using a color analysis tool, extract the facial skin color value of the second facial image;

Adjust the color component of the separated human face skin color area according to the human face skin color value to perform skin color adjustment;

Through the image reconstruction method, the skin color-adjusted human face skin color area is synthesized with other areas in the separated first human face image to obtain a third human face image.

Optionally, in a fourth implementation manner of the first aspect of the present invention, the skin color detection algorithm is used to determine the face skin color area of the first face image, and the color space of the face skin color area is Making the conversion includes:

Separate the foreground person and background in the first face image through an image segmentation algorithm;

Determine the facial skin color area of the foreground person through a skin color detection algorithm;

Convert the color space of the skin color area of the human face.

Optionally, in a fifth implementation manner of the first aspect of the present invention, obtaining the target face image through a trained generative adversarial network based on the third face image and the second face image includes: :

Train an initial generative adversarial network so that the trained generative adversarial network can learn the characteristics of human faces;

Extract facial features of the third facial image and facial features of the second facial image;

The facial features of the third facial image and the facial features of the second facial image are input into the generator of the generative adversarial network, and the generator converts the facial features of the third facial image into The facial features and the facial features of the second face image are fused to obtain the target face image.

Optionally, in a sixth implementation manner of the first aspect of the present invention, the skin color detection algorithm is used to determine the face skin color area of the first face image, and the color space of the face skin color area is Making the conversion includes:

Convert the first face image into a grayscale image;

Perform noise reduction and normalization processing on the grayscale image to obtain a preprocessed image;

Through the skin color detection algorithm, the human face skin color area of the preprocessed image is determined, and the color space of the human face skin color area is converted.

A second aspect of the present invention provides a face image fusion device, including: a memory and at least one processor, instructions are stored in the memory, and the memory and the at least one processor are interconnected through lines; the at least A processor calls the instruction in the memory to cause the face image fusion device to execute the above face image fusion method.

A third aspect of the present invention provides a computer-readable storage medium. The computer-readable storage medium stores instructions, which when run on a computer, cause the computer to execute the above-mentioned face image fusion method.

In the embodiment of the present invention, a first face image and a second face image are obtained; a skin color detection algorithm is used to determine the face skin color area of the first face image, and the color space of the face skin color area is Perform conversion; separate the face skin color area from the first face image according to the color space converted face skin color area and the preset color attribute range; extract the face of the second face image Skin color value: perform skin color adjustment on the separated human face skin color area according to the human face skin color color value, and compare the skin color adjusted human face skin color area with other areas in the separated first face image Synthesize to obtain a third face image; according to the third face image and the second face image, obtain a target face image through a trained generative adversarial network. In the present invention, first, the color space of the face skin color area of the first face image is converted, the face skin color color value of the second face image is extracted, and based on the face skin color color value, the separated face skin color area is Perform skin color adjustment, and then synthesize the adjusted facial skin color area with other areas in the separated first face image, and obtain the target face image through the trained generative adversarial network, which can be used before face image fusion. The skin color of the original face image is adjusted first, so that when faces with different skin colors are fused, the skin color blends more naturally, thereby improving the skin color fusion degree.

Description of drawings

Figure 1 is a schematic diagram of an embodiment of the face image fusion method in the embodiment of the present invention;

Figure 2 is a schematic diagram of an embodiment of a face image fusion device in an embodiment of the present invention;

Figure 3 is a schematic diagram of an embodiment of a face image fusion device in an embodiment of the present invention.

Detailed ways

Embodiments of the present invention provide a face image fusion method, device, equipment and storage medium.

Disclosed embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but rather that these embodiments are provided This is for a more thorough and complete understanding of this disclosure. It should be understood that the drawings and embodiments disclosed in the present invention are for illustrative purposes only and are not used to limit the scope of protection of the present invention.

In the description of the disclosed embodiments of the present invention, the term "including" and similar expressions shall be understood as an open inclusion, that is, "including but not limited to." The term "based on" should be understood to mean "based at least in part on." The terms "one embodiment" or "the embodiment" should be understood to mean "at least one embodiment". The terms "first," "second," etc. may refer to different or the same object. Other explicit and implicit definitions may be included below.

For ease of understanding, the specific process of the embodiment of the present invention is described below. Please refer to Figure 1. An embodiment of the face image fusion method in the embodiment of the present invention includes:

S100: Obtain the first face image and the second face image.

In this embodiment, the first face image is an unprocessed original face image, and the second face image is a reference face image for changes in the first face image. When fusing the face images, the first face image needs to be The face skin color of the image is adjusted to the face skin color of the second face image, and then the faces of the first face image and the second face image are exchanged.

S200: Determine the face skin color area of the first face image through the skin color detection algorithm, and convert the color space of the face skin color area.

In this embodiment, a skin color detection algorithm is used to determine the face skin color area of the first face image. The skin color detection algorithm is usually based on a color space model, such as RGB, HSV or YCbCr, etc.; the color space of the face skin color area is converted , wherein color space conversion refers to converting or representing color data in one color space into corresponding data in another color space, that is, using data in different color spaces to represent the same color. Optionally, in the present invention The RGB color space of the skin color area of the human face can be converted to the HSV color space. The HSV color space can better describe the sensory characteristics of the color, and it is more advantageous to identify colors on HSV than on RGB.

S300: Separate the human face skin color area from the first face image according to the human face skin color area after color space conversion and the preset color attribute range.

In this embodiment, by setting an appropriate color attribute range, the human face skin color area can be separated from the first face image. Specifically, the preset color attribute range of HSV is set. If the pixels of the human face skin color area are If the color attribute value is within the preset color attribute range, the pixel is marked, and then, based on all marked pixels, the face skin color area is separated from the first face image. In addition, a large amount of labeled image data can be used to train the model, and machine learning algorithms (such as support vector machines, neural networks, etc.) can be used to classify skin color and non-skin color areas.

S400: Extract the face skin color value of the second face image, perform skin color adjustment on the separated face skin color area according to the face skin color value, and compare the skin color adjusted face skin color area with the separated first face image The other areas in the image are synthesized to obtain the third face image.

In this embodiment, the facial skin color value of the second facial image is extracted, and the facial skin color value of the separated facial skin area is adjusted using the facial skin color value of the second facial image as a reference, for example, changing Brightness, saturation or hue, etc., for skin color adjustment, in which techniques such as color mapping, histogram equalization, and adjustment curves can be used to achieve skin color adjustment. After adjusting the skin color of the human face skin color area, the adjusted human face skin color area is synthesized or reconstructed with other areas in the separated first human face image to obtain a third human face image to maintain overall consistency and Naturalness, where the third face image refers to a new image formed by merging the skin color area of the face after skin color adjustment and the remaining areas in the first face image after the face skin color area is separated from the first face image. image.

500. According to the third face image and the second face image, obtain the target face image through the trained generative adversarial network.

In this embodiment, the third face image and the second face image are input into the trained generative adversarial network to obtain the target face image, where the generative adversarial network (GANs) is a type of network consisting of a generator and a discriminator. The network architecture of the generator attempts to generate realistic synthetic images, while the discriminator attempts to distinguish between real and synthetic images. Through continuous adversarial training, the generator and discriminator compete with each other and gradually improve the generator's ability to generate realistic images. The target face image is a new image obtained by face-changing the first face image.

In an optional implementation of the first aspect of the present invention, the preset color attribute range includes a preset hue range and a preset saturation range; according to the human face skin color area and the preset color attribute range after color space conversion, The face skin color area separated from the first face image includes:

Select the target pixel in the face skin color area after color space conversion according to the preset hue range and the preset saturation range; obtain the target face skin color area based on the target pixel; convert the target face skin color area from the first face image separated from it.

In this embodiment, in HSV mode, H (Hue) represents hue, S (Saturation) represents saturation, and B (Brightness) represents brightness. Skin color usually falls within a certain range of Hue and Saturation, where hue is at 0 It is measured by position on the ~360° standard color wheel. In common use, hue is identified by the color name, such as red, green or orange; saturation represents the purity of the color, when it is 0 it is gray, white , black and other gray colors have no saturation. At maximum saturation, each hue has the purest color light, ranging from 0 to 100%. In the present invention, it is first necessary to set the preset hue range and the preset saturation range, and then find the target pixels that meet the preset hue range and the preset saturation range among the pixels in the skin color area of the human face after color space conversion, Extract the target pixels to separate the skin color areas of the human face.

In an optional implementation of the first aspect of the present invention, selecting target pixels in the human face skin color area after color space conversion according to the preset hue range and the preset saturation range includes:

Generate a mask according to the preset hue range and the preset saturation range; perform bit operations on the mask and the first face image to obtain the target pixels in the skin color area of the face after color space conversion.

In this embodiment, you can set the hue range and saturation range available in HSV space through OpenCV, and then use the inRange function method to generate a mask, where the mask is a binary image; then use the bitwise_and() function to make this mask The code performs bit operations on the original image, that is, the first face image, to extract the target pixels in the skin color area of the face after color space conversion.

In an optional implementation of the first aspect of the present invention, the face skin color value of the second face image is extracted, the skin color is adjusted on the separated face skin color area according to the face skin color value, and the skin color is adjusted The facial skin color area is synthesized with other areas in the separated first face image to obtain the third face image including:

Use a color analysis tool to extract the face skin color value of the second face image; adjust the color component of the separated face skin color area according to the face skin color value to perform skin color adjustment; use the image reconstruction method to adjust the skin color The human face skin color area is synthesized with other areas in the separated first human face image to obtain a third human face image.

In this embodiment, by selecting an image with the required skin color, that is, the second face image, and using a color analysis tool, such as a color picker, to obtain the face skin color value in the image; the separated face Adjusting the color value of the skin color area to the required facial skin color value can be achieved by using color adjustment tools (such as color levels, curves, color balance, etc.) in image processing software. When adjusting the color value, you can refer to the required facial skin color value and gradually adjust the values of each color component (such as R, G, B) until the preset effect is achieved.

In an optional implementation of the first aspect of the present invention, determining the face skin color area of the first face image through a skin color detection algorithm, and converting the color space of the face skin color area includes:

The image segmentation algorithm is used to separate the foreground person and the background in the first face image; the skin color detection algorithm is used to determine the face skin color area of the foreground person; and the color space of the face skin color area is converted.

In this embodiment, an image segmentation algorithm is used to separate the foreground person and the background in the first face image. Commonly used image segmentation methods include threshold-based, edge detection-based, region growing-based, etc. For the extracted foreground character area, a skin color detection algorithm can be used to determine the face skin color area of the foreground character, and then convert the color space of the face skin color area.

In an optional implementation of the first aspect of the present invention, obtaining the target face image through a trained generative adversarial network based on the third face image and the second face image includes:

Train the initial generative adversarial network so that the trained generative adversarial network can learn the features of the face; extract the face features of the third face image and the face features of the second face image; combine the face features of the third face image , and the facial features of the second face image are input into the generator of the generative adversarial network. The generator fuses the facial features of the third face image and the facial features of the second face image to obtain the target face image. .

In this embodiment, in face image fusion, you first need to train a GAN model. Specifically, first initialize the parameters of the generator and discriminator networks; then extract n samples from the training set, and use the definition of the generator Generate n samples from the noise distribution; then fix the generator and train the discriminator to distinguish true from false as much as possible. After multiple update iterations, under ideal conditions, the final discriminator cannot distinguish whether the picture comes from the real training sample set. Just use the samples generated by the generator. At this time, the probability of discrimination is 0.5, and the training is completed. The trained GAN model can learn the features of the face. Then, by inputting the face features of the third face image and the face features of the second face image into the generator, the generator will try to combine the two people. The facial features are synthesized into a new face image, and the generated image will have the identity characteristics of one person, but at the same time reflect the appearance characteristics of another person.

In an optional implementation of the first aspect of the present invention, determining the face skin color area of the first face image through a skin color detection algorithm, and converting the color space of the face skin color area includes:

Convert the first face image into a grayscale image; perform noise reduction and normalization processing on the grayscale image to obtain a preprocessed image; determine the facial skin color area of the preprocessed image through the skin color detection algorithm, and compare the facial skin color Convert to the regional color space.

In this embodiment, before performing face image fusion, the first face image needs to be preprocessed. The first face image is grayscaled to obtain a grayscale image. The color image usually includes R, G The three components of R, G and B respectively display various colors such as red, green and blue. Grayscale is the process of making the three components of R, G and B of the color image equal. Then, the grayscale image is denoised and normalized to eliminate irrelevant information in the original image and restore useful real information, thereby improving the reliability of image fusion.

Referring to Figure 2, a second aspect of the present invention provides a face image fusion device. The face image fusion device includes:

The image acquisition module 10 is used to acquire a first face image and a second face image. The first face image is an unprocessed original face image, and the second face image is a reference for the change of the first face image. face image;

The color space conversion module 20 is used to determine the facial skin color area of the first face image through the skin color detection algorithm, and convert the color space of the human face skin color area;

The image separation module 30 is used to separate the human face skin color area from the first human face image according to the human face skin color area after color space conversion and the preset color attribute range;

The image synthesis module 40 is used to extract the face skin color value of the second face image, perform skin color adjustment on the separated face skin color area according to the face skin color value, and combine the skin color adjusted face skin color area with the separated face skin color area. The other areas in the first face image are synthesized to obtain the third face image;

The target image acquisition module 50 is used to obtain the target face image through the trained generative adversarial network based on the third face image and the second face image.

In an optional implementation of the second aspect of the present invention, the image separation module 30 is also used to select target pixels in the human face skin color area after color space conversion according to the preset hue range and the preset saturation range; according to Target pixels are used to obtain the target face skin color area; the target face skin color area is separated from the first face image.

In an optional implementation of the second aspect of the present invention, the image separation module 30 is also used to generate a mask according to the preset hue range and the preset saturation range; and perform bit operations on the mask and the first face image. , obtain the target pixels in the skin color area of the human face after color space conversion.

In an optional implementation of the second aspect of the present invention, the image synthesis module 40 is also used to extract the facial skin color value of the second facial image through a color analysis tool; adjust the separated facial skin color value according to the facial skin color value. The color component of the human face skin color area is used for skin color adjustment; through the image reconstruction method, the skin color adjusted human face skin color area is synthesized with other areas in the separated first face image to obtain a third face image.

In an optional implementation of the second aspect of the present invention, the color space conversion module 20 is also used to separate the foreground person from the background in the first face image through an image segmentation algorithm; and determine the foreground person through a skin color detection algorithm. The human face skin color area in; convert the color space of the human face skin color area.

In an optional implementation of the second aspect of the present invention, the target image acquisition module 50 is also used to train the initial generative adversarial network, so that the trained generative adversarial network can learn the characteristics of the human face; extract the person in the third face image facial features and the facial features of the second face image; input the facial features of the third face image and the facial features of the second face image into the generator of the generative adversarial network, and the generator will The facial features of the face image and the facial features of the second face image are fused to obtain the target face image.

In an optional implementation of the second aspect of the present invention, the color space conversion module 20 is also used to convert the first face image into a grayscale image; perform noise reduction and normalization processing on the grayscale image to obtain a predetermined Process the image; use the skin color detection algorithm to determine the facial skin color area of the pre-processed image, and convert the color space of the human face skin color area.

Figure 3 is a schematic structural diagram of a face image fusion device provided by an embodiment of the present invention. The face image fusion device 500 may vary greatly due to different configurations or performance, and may include one or more central processors (central processors). processing units (CPU) 510 (eg, one or more processors) and memory 520 , one or more storage media 530 (eg, one or more mass storage devices) that stores applications 533 or data 532 . Among them, the memory 520 and the storage medium 530 may be short-term storage or persistent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the face image fusion device 500 . Furthermore, the processor 510 may be configured to communicate with the storage medium 530 and execute a series of instruction operations in the storage medium 530 on the face image fusion device 500 .

The facial image fusion device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and/or one or more operating systems 531, for example Windows Server, Mac OS X, Unix, Linux, Free BSD and many more. Those skilled in the art can understand that the structure of the facial image fusion device shown in Figure 3 does not constitute a limitation on the facial image fusion device, and may include more or fewer components than shown in the figure, or combine certain components. Or a different component arrangement.

The present invention also provides a computer-readable storage medium. The computer-readable storage medium can be a non-volatile computer-readable storage medium. The computer-readable storage medium can also be a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium, and when the instructions are run on the computer, they cause the computer to execute the steps of the face image fusion method.

In the context of this disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media may include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media would include one or more wire-based electrical connections, laptop disks, hard drives, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.

Furthermore, although operations are depicted in a specific order, this should be understood to require that such operations be performed in the specific order shown or in sequential order, or that all illustrated operations should be performed to achieve desirable results. Under certain circumstances, multitasking and parallel processing may be advantageous. Likewise, although several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the present disclosure. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation can also be implemented in multiple implementations separately or in any suitable subcombination.

Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are merely example forms of implementing the claims.

Claims (10)

1. The face image fusion method is characterized by comprising the following steps of:

acquiring a first face image and a second face image, wherein the first face image is an unprocessed original face image, and the second face image is a face image referenced by the change of the first face image;

determining a human face skin color region of the first human face image through a skin color detection algorithm, and converting the color space of the human face skin color region;

separating the human face skin color region from the first human face image according to the human face skin color region after the color space conversion and a preset color attribute range;

extracting a human face skin color value of the second human face image, carrying out skin color adjustment on the separated human face skin color region according to the human face skin color value, and synthesizing the human face skin color region with the skin color adjusted with other regions in the first human face image after separation to obtain a third human face image;

and obtaining a target face image through a training generation countermeasure network according to the third face image and the second face image.

2. The face image fusion method of claim 1, wherein the predetermined color attribute range includes a predetermined hue range and a predetermined saturation range;

the separating the face skin color region from the first face image according to the face skin color region after the color space conversion and the preset color attribute range comprises the following steps:

selecting target pixels in the face skin color region after the color space conversion according to the preset hue range and the preset saturation range;

obtaining a target face skin color region according to the target pixel;

and separating the target face skin color area from the first face image.

3. The method according to claim 2, wherein selecting the target pixel in the color space converted face skin region according to the preset hue range and the preset saturation range comprises:

generating a mask according to the preset tone range and the preset saturation range;

and carrying out bit operation on the mask and the first face image to obtain target pixels in the face skin region after the color space conversion.

4. The method of claim 1, wherein the extracting the face skin color value of the second face image, adjusting the skin color of the separated face skin color region according to the face skin color value, and combining the face skin color region with other regions in the first face image after the skin color adjustment, to obtain a third face image comprises:

extracting a face skin color value of the second face image through a color analysis tool;

according to the skin color value of the human face, adjusting the color component of the separated skin color region of the human face so as to adjust the skin color;

and synthesizing the human face skin color region with the skin color adjusted with other regions in the first human face image after separation by an image reconstruction method to obtain a third human face image.

5. The method according to claim 1, wherein the determining a face skin color region of the first face image by a skin color detection algorithm and converting a color space of the face skin color region comprises:

separating a foreground person from a background in the first face image through an image segmentation algorithm;

determining a face skin color region in the foreground person through a skin color detection algorithm;

and converting the color space of the human face skin color region.

6. The face image fusion method of claim 1, wherein the obtaining the target face image through the training generation countermeasure network according to the third face image and the second face image comprises:

training an initial generation countermeasure network, so that the training generation countermeasure network can learn the characteristics of the face;

extracting the face features of the third face image and the face features of the second face image;

inputting the face features of the third face image and the face features of the second face image into the generator for generating the countermeasure network, and fusing the face features of the third face image and the face features of the second face image by the generator to obtain a target face image.

7. The method according to claim 1, wherein the determining a face skin color region of the first face image by a skin color detection algorithm and converting a color space of the face skin color region comprises:

converting the first face image into a gray scale image;

carrying out noise reduction and normalization treatment on the gray level image to obtain a preprocessed image;

and determining a human face skin color region of the preprocessed image through a skin color detection algorithm, and converting the color space of the human face skin color region.

8. A face image fusion apparatus, characterized in that the face image fusion apparatus comprises:

the image acquisition module is used for acquiring a first face image and a second face image, wherein the first face image is an unprocessed original face image, and the second face image is a face image referenced by the change of the first face image;

the color space conversion module is used for determining a human face skin color region of the first human face image through a skin color detection algorithm and converting the color space of the human face skin color region;

the image separation module is used for separating the human face skin color region from the first human face image according to the human face skin color region after the color space conversion and a preset color attribute range;

the image synthesis module is used for extracting a human face skin color value of the second human face image, carrying out skin color adjustment on the separated human face skin color region according to the human face skin color value, and synthesizing the human face skin color region with the skin color adjusted with other regions in the first human face image after separation to obtain a third human face image;

and the target image acquisition module is used for acquiring a target face image through a training generation countermeasure network according to the third face image and the second face image.

9. A face image fusion apparatus, characterized in that the face image fusion apparatus comprises: a memory and at least one processor, the memory having instructions stored therein, the memory and the at least one processor being interconnected by a line;

the at least one processor invokes the instructions in the memory to cause the face image fusion apparatus to perform the face image fusion method of any one of claims 1-7.

10. A computer readable storage medium having a computer program stored thereon, wherein the computer program when executed by a processor implements the face image fusion method according to any of claims 1-7.